* [net-next 01/14] i40evf: Use net_device_stats from struct net_device
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 02/14] i40e/i40evf: Remove VF Rx csum offload for tunneled packets Jeff Kirsher
` (13 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Tobias Klauser, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Tobias Klauser <tklauser@distanz.ch>
Instead of using a private copy of struct net_device_stats in
struct i40evf_adapter, use stats from struct net_device. Also remove the
now unnecessary .ndo_get_stats function.
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40evf/i40evf.h | 1 -
drivers/net/ethernet/intel/i40evf/i40evf_main.c | 16 ----------------
.../net/ethernet/intel/i40evf/i40evf_virtchnl.c | 22 +++++++++++-----------
3 files changed, 11 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf.h b/drivers/net/ethernet/intel/i40evf/i40evf.h
index 35ded19e9cc2..7bbd11abed08 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf.h
+++ b/drivers/net/ethernet/intel/i40evf/i40evf.h
@@ -252,7 +252,6 @@ struct i40evf_adapter {
/* OS defined structs */
struct net_device *netdev;
struct pci_dev *pdev;
- struct net_device_stats net_stats;
struct i40e_hw hw; /* defined in i40e_type.h */
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_main.c b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
index 12a930e879af..671913c74bf8 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c
+++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
@@ -2243,21 +2243,6 @@ static int i40evf_close(struct net_device *netdev)
}
/**
- * i40evf_get_stats - Get System Network Statistics
- * @netdev: network interface device structure
- *
- * Returns the address of the device statistics structure.
- * The statistics are actually updated from the timer callback.
- **/
-static struct net_device_stats *i40evf_get_stats(struct net_device *netdev)
-{
- struct i40evf_adapter *adapter = netdev_priv(netdev);
-
- /* only return the current stats */
- return &adapter->net_stats;
-}
-
-/**
* i40evf_change_mtu - Change the Maximum Transfer Unit
* @netdev: network interface device structure
* @new_mtu: new value for maximum frame size
@@ -2363,7 +2348,6 @@ static const struct net_device_ops i40evf_netdev_ops = {
.ndo_open = i40evf_open,
.ndo_stop = i40evf_close,
.ndo_start_xmit = i40evf_xmit_frame,
- .ndo_get_stats = i40evf_get_stats,
.ndo_set_rx_mode = i40evf_set_rx_mode,
.ndo_validate_addr = eth_validate_addr,
.ndo_set_mac_address = i40evf_set_mac,
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_virtchnl.c b/drivers/net/ethernet/intel/i40evf/i40evf_virtchnl.c
index 3bccfbb1db14..deb2cb8dac6b 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf_virtchnl.c
+++ b/drivers/net/ethernet/intel/i40evf/i40evf_virtchnl.c
@@ -960,17 +960,17 @@ void i40evf_virtchnl_completion(struct i40evf_adapter *adapter,
case I40E_VIRTCHNL_OP_GET_STATS: {
struct i40e_eth_stats *stats =
(struct i40e_eth_stats *)msg;
- adapter->net_stats.rx_packets = stats->rx_unicast +
- stats->rx_multicast +
- stats->rx_broadcast;
- adapter->net_stats.tx_packets = stats->tx_unicast +
- stats->tx_multicast +
- stats->tx_broadcast;
- adapter->net_stats.rx_bytes = stats->rx_bytes;
- adapter->net_stats.tx_bytes = stats->tx_bytes;
- adapter->net_stats.tx_errors = stats->tx_errors;
- adapter->net_stats.rx_dropped = stats->rx_discards;
- adapter->net_stats.tx_dropped = stats->tx_discards;
+ netdev->stats.rx_packets = stats->rx_unicast +
+ stats->rx_multicast +
+ stats->rx_broadcast;
+ netdev->stats.tx_packets = stats->tx_unicast +
+ stats->tx_multicast +
+ stats->tx_broadcast;
+ netdev->stats.rx_bytes = stats->rx_bytes;
+ netdev->stats.tx_bytes = stats->tx_bytes;
+ netdev->stats.tx_errors = stats->tx_errors;
+ netdev->stats.rx_dropped = stats->rx_discards;
+ netdev->stats.tx_dropped = stats->tx_discards;
adapter->current_stats = *stats;
}
break;
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 02/14] i40e/i40evf: Remove VF Rx csum offload for tunneled packets
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
2017-04-20 1:57 ` [net-next 01/14] i40evf: Use net_device_stats from struct net_device Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 03/14] i40e: Fix support for flow director programming status Jeff Kirsher
` (12 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem
Cc: alice michael, netdev, nhorman, sassmann, jogreene,
Preethi Banala, Jeff Kirsher
From: alice michael <alice.michael@intel.com>
Rx checksum offload for tunneled packets was never being negotiated or
requested by VF. This capability was assumed by default and enabled in
current hardware for VF. Going forward, this feature needs to be disabled
or advanced ptypes should be negotiated with PF in the future.
Change-ID: I9e54cfa8a90e03ab6956db4412f1e337ccd2c2e0
Signed-off-by: Preethi Banala <preethi.banala@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40evf/i40e_txrx.c | 7 -------
1 file changed, 7 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
index 460171edc412..fc4b14af370d 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
@@ -831,13 +831,6 @@ static inline void i40e_rx_checksum(struct i40e_vsi *vsi,
if (rx_error & BIT(I40E_RX_DESC_ERROR_PPRS_SHIFT))
return;
- /* If there is an outer header present that might contain a checksum
- * we need to bump the checksum level by 1 to reflect the fact that
- * we are indicating we validated the inner checksum.
- */
- if (decoded.tunnel_type >= I40E_RX_PTYPE_TUNNEL_IP_GRENAT)
- skb->csum_level = 1;
-
/* Only report checksum unnecessary for TCP, UDP, or SCTP */
switch (decoded.inner_prot) {
case I40E_RX_PTYPE_INNER_PROT_TCP:
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 03/14] i40e: Fix support for flow director programming status
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
2017-04-20 1:57 ` [net-next 01/14] i40evf: Use net_device_stats from struct net_device Jeff Kirsher
2017-04-20 1:57 ` [net-next 02/14] i40e/i40evf: Remove VF Rx csum offload for tunneled packets Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 04/14] i40e: dump VF information in debugfs Jeff Kirsher
` (11 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Alexander Duyck, netdev, nhorman, sassmann, jogreene,
Jeff Kirsher
From: Alexander Duyck <alexander.h.duyck@intel.com>
This patch fixes an issue I introduced when I converted the code over to
using the length field to determine if a descriptor was done or not. It
turns out that we are also processing programming descriptors in the Rx
path and need to have these processed even though the length field will be
0 on these packets. What will happen with a programming descriptor is that
we will receive a descriptor that has the SPH bit set, and the header
length and packet length fields cleared.
To account for this we should be checking for the bit for split header
being set even though we aren't actually using header split. This bit is
set in the length field to indicate if a programming descriptor response is
contained in the descriptor. Since we don't support header split we don't
need to perform the extra checks of using a fixed value for the entire
length field.
In addition I am moving the function for checking if a filter is a
programming status filter into the i40e_txrx.c file since there is no
longer support for FCoE it doesn't make sense to keep this file in i40e.h.
Change-ID: I12c359c3dc70adb9d6b92b27324bb2c7f04c1a06
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e.h | 15 --------
drivers/net/ethernet/intel/i40e/i40e_txrx.c | 50 ++++++++++++++++++++-------
drivers/net/ethernet/intel/i40evf/i40e_txrx.c | 9 ++---
3 files changed, 43 insertions(+), 31 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index e987503f8517..b629f3adcecb 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -763,21 +763,6 @@ static inline void i40e_vsi_setup_irqhandler(struct i40e_vsi *vsi,
}
/**
- * i40e_rx_is_programming_status - check for programming status descriptor
- * @qw: the first quad word of the program status descriptor
- *
- * The value of in the descriptor length field indicate if this
- * is a programming status descriptor for flow director or FCoE
- * by the value of I40E_RX_PROG_STATUS_DESC_LENGTH, otherwise
- * it is a packet descriptor.
- **/
-static inline bool i40e_rx_is_programming_status(u64 qw)
-{
- return I40E_RX_PROG_STATUS_DESC_LENGTH ==
- (qw >> I40E_RX_PROG_STATUS_DESC_LENGTH_SHIFT);
-}
-
-/**
* i40e_get_fd_cnt_all - get the total FD filter space available
* @pf: pointer to the PF struct
**/
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index 20691d2bf113..843d7ffe1784 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -1038,9 +1038,29 @@ static bool i40e_set_new_dynamic_itr(struct i40e_ring_container *rc)
}
/**
+ * i40e_rx_is_programming_status - check for programming status descriptor
+ * @qw: qword representing status_error_len in CPU ordering
+ *
+ * The value of in the descriptor length field indicate if this
+ * is a programming status descriptor for flow director or FCoE
+ * by the value of I40E_RX_PROG_STATUS_DESC_LENGTH, otherwise
+ * it is a packet descriptor.
+ **/
+static inline bool i40e_rx_is_programming_status(u64 qw)
+{
+ /* The Rx filter programming status and SPH bit occupy the same
+ * spot in the descriptor. Since we don't support packet split we
+ * can just reuse the bit as an indication that this is a
+ * programming status descriptor.
+ */
+ return qw & I40E_RXD_QW1_LENGTH_SPH_MASK;
+}
+
+/**
* i40e_clean_programming_status - clean the programming status descriptor
* @rx_ring: the rx ring that has this descriptor
* @rx_desc: the rx descriptor written back by HW
+ * @qw: qword representing status_error_len in CPU ordering
*
* Flow director should handle FD_FILTER_STATUS to check its filter programming
* status being successful or not and take actions accordingly. FCoE should
@@ -1048,12 +1068,18 @@ static bool i40e_set_new_dynamic_itr(struct i40e_ring_container *rc)
*
**/
static void i40e_clean_programming_status(struct i40e_ring *rx_ring,
- union i40e_rx_desc *rx_desc)
+ union i40e_rx_desc *rx_desc,
+ u64 qw)
{
- u64 qw;
+ u32 ntc = rx_ring->next_to_clean + 1;
u8 id;
- qw = le64_to_cpu(rx_desc->wb.qword1.status_error_len);
+ /* fetch, update, and store next to clean */
+ ntc = (ntc < rx_ring->count) ? ntc : 0;
+ rx_ring->next_to_clean = ntc;
+
+ prefetch(I40E_RX_DESC(rx_ring, ntc));
+
id = (qw & I40E_RX_PROG_STATUS_DESC_QW1_PROGID_MASK) >>
I40E_RX_PROG_STATUS_DESC_QW1_PROGID_SHIFT;
@@ -1911,11 +1937,6 @@ static bool i40e_is_non_eop(struct i40e_ring *rx_ring,
prefetch(I40E_RX_DESC(rx_ring, ntc));
-#define staterrlen rx_desc->wb.qword1.status_error_len
- if (unlikely(i40e_rx_is_programming_status(le64_to_cpu(staterrlen)))) {
- i40e_clean_programming_status(rx_ring, rx_desc);
- return true;
- }
/* if we are the last buffer then there is nothing else to do */
#define I40E_RXD_EOF BIT(I40E_RX_DESC_STATUS_EOF_SHIFT)
if (likely(i40e_test_staterr(rx_desc, I40E_RXD_EOF)))
@@ -1968,10 +1989,6 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
* hardware wrote DD then the length will be non-zero
*/
qword = le64_to_cpu(rx_desc->wb.qword1.status_error_len);
- size = (qword & I40E_RXD_QW1_LENGTH_PBUF_MASK) >>
- I40E_RXD_QW1_LENGTH_PBUF_SHIFT;
- if (!size)
- break;
/* This memory barrier is needed to keep us from reading
* any other fields out of the rx_desc until we have
@@ -1979,6 +1996,15 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
*/
dma_rmb();
+ if (unlikely(i40e_rx_is_programming_status(qword))) {
+ i40e_clean_programming_status(rx_ring, rx_desc, qword);
+ continue;
+ }
+ size = (qword & I40E_RXD_QW1_LENGTH_PBUF_MASK) >>
+ I40E_RXD_QW1_LENGTH_PBUF_SHIFT;
+ if (!size)
+ break;
+
rx_buffer = i40e_get_rx_buffer(rx_ring, size);
/* retrieve a buffer from the ring */
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
index fc4b14af370d..80931e3f9445 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
@@ -1317,10 +1317,6 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
* hardware wrote DD then the length will be non-zero
*/
qword = le64_to_cpu(rx_desc->wb.qword1.status_error_len);
- size = (qword & I40E_RXD_QW1_LENGTH_PBUF_MASK) >>
- I40E_RXD_QW1_LENGTH_PBUF_SHIFT;
- if (!size)
- break;
/* This memory barrier is needed to keep us from reading
* any other fields out of the rx_desc until we have
@@ -1328,6 +1324,11 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
*/
dma_rmb();
+ size = (qword & I40E_RXD_QW1_LENGTH_PBUF_MASK) >>
+ I40E_RXD_QW1_LENGTH_PBUF_SHIFT;
+ if (!size)
+ break;
+
rx_buffer = i40e_get_rx_buffer(rx_ring, size);
/* retrieve a buffer from the ring */
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (2 preceding siblings ...)
2017-04-20 1:57 ` [net-next 03/14] i40e: Fix support for flow director programming status Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 6:16 ` Or Gerlitz
2017-04-20 1:57 ` [net-next 05/14] i40e/i40evf: Add tracepoints Jeff Kirsher
` (10 subsequent siblings)
14 siblings, 1 reply; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Mitch Williams, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Mitch Williams <mitch.a.williams@intel.com>
Dump some internal state about VFs through debugfs. This provides
information not available with 'ip link show'. To use, write "dump vf
<id>" to the command file, or just "dump vf" to dump information on all
of the VFs.
Change-ID: Ibe32b7f4ae55d4358c0b903217475f708ada1ecd
Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e_debugfs.c | 51 ++++++++++++++++++++++++++
1 file changed, 51 insertions(+)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_debugfs.c b/drivers/net/ethernet/intel/i40e/i40e_debugfs.c
index c5f68cc1edcd..a3d7ec62b76c 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_debugfs.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_debugfs.c
@@ -384,6 +384,8 @@ static void i40e_dbg_dump_vsi_seid(struct i40e_pf *pf, int seid)
" base_queue = %d, num_queue_pairs = %d, num_desc = %d\n",
vsi->base_queue, vsi->num_queue_pairs, vsi->num_desc);
dev_info(&pf->pdev->dev, " type = %i\n", vsi->type);
+ if (vsi->type == I40E_VSI_SRIOV)
+ dev_info(&pf->pdev->dev, " VF ID = %i\n", vsi->vf_id);
dev_info(&pf->pdev->dev,
" info: valid_sections = 0x%04x, switch_id = 0x%04x\n",
vsi->info.valid_sections, vsi->info.switch_id);
@@ -694,6 +696,47 @@ static void i40e_dbg_dump_veb_all(struct i40e_pf *pf)
}
}
+/**
+ * i40e_dbg_dump_vf - dump VF info
+ * @pf: the i40e_pf created in command write
+ * @vf_id: the vf_id from the user
+ **/
+static void i40e_dbg_dump_vf(struct i40e_pf *pf, int vf_id)
+{
+ struct i40e_vf *vf;
+ struct i40e_vsi *vsi;
+
+ if (!pf->num_alloc_vfs) {
+ dev_info(&pf->pdev->dev, "no VFs allocated\n");
+ } else if ((vf_id >= 0) && (vf_id < pf->num_alloc_vfs)) {
+ vf = &pf->vf[vf_id];
+ vsi = pf->vsi[vf->lan_vsi_idx];
+ dev_info(&pf->pdev->dev, "vf %2d: VSI id=%d, seid=%d, qps=%d\n",
+ vf_id, vf->lan_vsi_id, vsi->seid, vf->num_queue_pairs);
+ dev_info(&pf->pdev->dev, " num MDD=%lld, invalid msg=%lld, valid msg=%lld\n",
+ vf->num_mdd_events,
+ vf->num_invalid_msgs,
+ vf->num_valid_msgs);
+ } else {
+ dev_info(&pf->pdev->dev, "invalid VF id %d\n", vf_id);
+ }
+}
+
+/**
+ * i40e_dbg_dump_vf_all - dump VF info for all VFs
+ * @pf: the i40e_pf created in command write
+ **/
+static void i40e_dbg_dump_vf_all(struct i40e_pf *pf)
+{
+ int i;
+
+ if (!pf->num_alloc_vfs)
+ dev_info(&pf->pdev->dev, "no VFs enabled!\n");
+ else
+ for (i = 0; i < pf->num_alloc_vfs; i++)
+ i40e_dbg_dump_vf(pf, i);
+}
+
#define I40E_MAX_DEBUG_OUT_BUFFER (4096*4)
/**
* i40e_dbg_command_write - write into command datum
@@ -712,6 +755,7 @@ static ssize_t i40e_dbg_command_write(struct file *filp,
struct i40e_vsi *vsi;
int vsi_seid;
int veb_seid;
+ int vf_id;
int cnt;
/* don't allow partial writes */
@@ -914,6 +958,12 @@ static ssize_t i40e_dbg_command_write(struct file *filp,
i40e_dbg_dump_veb_seid(pf, vsi_seid);
else
i40e_dbg_dump_veb_all(pf);
+ } else if (strncmp(&cmd_buf[5], "vf", 2) == 0) {
+ cnt = sscanf(&cmd_buf[7], "%i", &vf_id);
+ if (cnt > 0)
+ i40e_dbg_dump_vf(pf, vf_id);
+ else
+ i40e_dbg_dump_vf_all(pf);
} else if (strncmp(&cmd_buf[5], "desc", 4) == 0) {
int ring_id, desc_n;
if (strncmp(&cmd_buf[10], "rx", 2) == 0) {
@@ -1109,6 +1159,7 @@ static ssize_t i40e_dbg_command_write(struct file *filp,
dev_info(&pf->pdev->dev, "dump vsi [seid]\n");
dev_info(&pf->pdev->dev, "dump reset stats\n");
dev_info(&pf->pdev->dev, "dump port\n");
+ dev_info(&pf->pdev->dev, "dump vf [vf_id]\n");
dev_info(&pf->pdev->dev,
"dump debug fwdata <cluster_id> <table_id> <index>\n");
}
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 1:57 ` [net-next 04/14] i40e: dump VF information in debugfs Jeff Kirsher
@ 2017-04-20 6:16 ` Or Gerlitz
2017-04-20 10:09 ` Mintz, Yuval
0 siblings, 1 reply; 31+ messages in thread
From: Or Gerlitz @ 2017-04-20 6:16 UTC (permalink / raw)
To: Jeff Kirsher, Mitch Williams
Cc: David Miller, Linux Netdev List, nhorman@redhat.com,
sassmann@redhat.com, jogreene@redhat.com
On Thu, Apr 20, 2017 at 4:57 AM, Jeff Kirsher
<jeffrey.t.kirsher@intel.com> wrote:
> From: Mitch Williams <mitch.a.williams@intel.com>
>
> Dump some internal state about VFs through debugfs. This provides
> information not available with 'ip link show'.
such as?
donnwantobethedebugfspolice, but still, in the 2-3 times we tried to
push debugfs to MLNX NIC
drivers, Dave disallowed that, and lately the switch team even went
further and deleted that
portion of the mlxsw driver -- all to all, I don't see much point for
these type of changes, thoughts?
^ permalink raw reply [flat|nested] 31+ messages in thread
* RE: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 6:16 ` Or Gerlitz
@ 2017-04-20 10:09 ` Mintz, Yuval
2017-04-20 11:39 ` Jiri Pirko
2017-04-20 15:20 ` David Miller
0 siblings, 2 replies; 31+ messages in thread
From: Mintz, Yuval @ 2017-04-20 10:09 UTC (permalink / raw)
To: Or Gerlitz, Jeff Kirsher, Mitch Williams, David Miller
Cc: Linux Netdev List, nhorman@redhat.com, sassmann@redhat.com,
jogreene@redhat.com
> > Dump some internal state about VFs through debugfs. This provides
> > information not available with 'ip link show'.
>
> such as?
>
> donnwantobethedebugfspolice, but still, in the 2-3 times we tried to push
> debugfs to MLNX NIC drivers, Dave disallowed that, and lately the switch
> team even went further and deleted that portion of the mlxsw driver -- all to
> all, I don't see much point for these type of changes, thoughts?
Don't want to hikjack your thread, but continuing this topic -
Is there some flat-out disapproval for debugfs in net-next now?
We're currently internally engaged with adding qed support for register dumps
[~equivalents for `ethtool -d' outputs] through debugfs, on behalf of storage
drivers [qedi/qedf] lacking the API for doing that.
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 10:09 ` Mintz, Yuval
@ 2017-04-20 11:39 ` Jiri Pirko
2017-04-20 13:57 ` Mintz, Yuval
2017-04-20 15:20 ` David Miller
1 sibling, 1 reply; 31+ messages in thread
From: Jiri Pirko @ 2017-04-20 11:39 UTC (permalink / raw)
To: Mintz, Yuval
Cc: Or Gerlitz, Jeff Kirsher, Mitch Williams, David Miller,
Linux Netdev List, nhorman@redhat.com, sassmann@redhat.com,
jogreene@redhat.com
Thu, Apr 20, 2017 at 12:09:28PM CEST, Yuval.Mintz@cavium.com wrote:
>> > Dump some internal state about VFs through debugfs. This provides
>> > information not available with 'ip link show'.
>>
>> such as?
>>
>> donnwantobethedebugfspolice, but still, in the 2-3 times we tried to push
>> debugfs to MLNX NIC drivers, Dave disallowed that, and lately the switch
>> team even went further and deleted that portion of the mlxsw driver -- all to
>> all, I don't see much point for these type of changes, thoughts?
>
>Don't want to hikjack your thread, but continuing this topic -
>Is there some flat-out disapproval for debugfs in net-next now?
It was mentioned by DaveM on the list multiple times.
>
>We're currently internally engaged with adding qed support for register dumps
>[~equivalents for `ethtool -d' outputs] through debugfs, on behalf of storage
>drivers [qedi/qedf] lacking the API for doing that.
That sounds wrong. Either introduce a generic infra to expose the info you
need or you don't do that. For your inhouse debugging, you should have
oot patch to expose whatever you need.
^ permalink raw reply [flat|nested] 31+ messages in thread
* RE: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 11:39 ` Jiri Pirko
@ 2017-04-20 13:57 ` Mintz, Yuval
2017-04-20 15:51 ` David Miller
0 siblings, 1 reply; 31+ messages in thread
From: Mintz, Yuval @ 2017-04-20 13:57 UTC (permalink / raw)
To: Jiri Pirko
Cc: Or Gerlitz, Jeff Kirsher, Mitch Williams, David Miller,
Linux Netdev List, nhorman@redhat.com, sassmann@redhat.com,
jogreene@redhat.com
> >> > Dump some internal state about VFs through debugfs. This provides
> >> > information not available with 'ip link show'.
> >>
> >> such as?
> >>
> >> donnwantobethedebugfspolice, but still, in the 2-3 times we tried to
> >> push debugfs to MLNX NIC drivers, Dave disallowed that, and lately
> >> the switch team even went further and deleted that portion of the
> >> mlxsw driver -- all to all, I don't see much point for these type of changes,
> thoughts?
> >
> >Don't want to hikjack your thread, but continuing this topic - Is there
> >some flat-out disapproval for debugfs in net-next now?
>
> It was mentioned by DaveM on the list multiple times.
> >
> >We're currently internally engaged with adding qed support for register
> >dumps [~equivalents for `ethtool -d' outputs] through debugfs, on
> >behalf of storage drivers [qedi/qedf] lacking the API for doing that.
>
> That sounds wrong. Either introduce a generic infra to expose the info you
> need or you don't do that.
I agree this surely isn't *our* convention, but for scsi using debugfs
[or sysfs] is a common practice for debug purposes.
Also, our HW debug capabilities are highly-customable, and I want to
have the ability to configure those on the fly [E.g., dynamically
configuring various HW events to be recorded].
Each such configuration involves multiple register writes and reads
according to user provided inputs.
I don't really see how to generalize the information collection in a way
that would benefit anyone else.
> For your inhouse debugging, you should have oot
> patch to expose whatever you need.
I don't want in-house debugging capabilities -
I want field debug capabilities.
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 13:57 ` Mintz, Yuval
@ 2017-04-20 15:51 ` David Miller
2017-04-20 16:24 ` Mintz, Yuval
0 siblings, 1 reply; 31+ messages in thread
From: David Miller @ 2017-04-20 15:51 UTC (permalink / raw)
To: Yuval.Mintz
Cc: jiri, gerlitz.or, jeffrey.t.kirsher, mitch.a.williams, netdev,
nhorman, sassmann, jogreene
From: "Mintz, Yuval" <Yuval.Mintz@cavium.com>
Date: Thu, 20 Apr 2017 13:57:08 +0000
> I agree this surely isn't *our* convention, but for scsi using debugfs
> [or sysfs] is a common practice for debug purposes.
>
> Also, our HW debug capabilities are highly-customable, and I want to
> have the ability to configure those on the fly [E.g., dynamically
> configuring various HW events to be recorded].
> Each such configuration involves multiple register writes and reads
> according to user provided inputs.
> I don't really see how to generalize the information collection in a way
> that would benefit anyone else.
That's basically what everyone says who slaps random crap into debugfs.
>> For your inhouse debugging, you should have oot
>> patch to expose whatever you need.
>
> I don't want in-house debugging capabilities -
> I want field debug capabilities.
Then it has to use a portable, well defined, set of interfaces.
^ permalink raw reply [flat|nested] 31+ messages in thread
* RE: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 15:51 ` David Miller
@ 2017-04-20 16:24 ` Mintz, Yuval
2017-04-20 16:58 ` David Miller
0 siblings, 1 reply; 31+ messages in thread
From: Mintz, Yuval @ 2017-04-20 16:24 UTC (permalink / raw)
To: David Miller
Cc: jiri@resnulli.us, gerlitz.or@gmail.com,
jeffrey.t.kirsher@intel.com, mitch.a.williams@intel.com,
netdev@vger.kernel.org, nhorman@redhat.com, sassmann@redhat.com,
jogreene@redhat.com, Dupuis, Chad, Rangankar, Manish
> > I agree this surely isn't *our* convention, but for scsi using
> > debugfs [or sysfs] is a common practice for debug purposes.
> >
> > Also, our HW debug capabilities are highly-customable, and I want to
> > have the ability to configure those on the fly [E.g., dynamically
> > configuring various HW events to be recorded].
> > Each such configuration involves multiple register writes and reads
> > according to user provided inputs.
> > I don't really see how to generalize the information collection in a
> > way that would benefit anyone else.
>
> That's basically what everyone says who slaps random crap into debugfs.
True; But assuming that means the suggestion to use debugfs means the
content is random crap is a logical fallacy.
> >> For your inhouse debugging, you should have oot patch to expose
> >> whatever you need.
> >
> > I don't want in-house debugging capabilities - I want field debug
> > capabilities.
>
> Then it has to use a portable, well defined, set of interfaces.
How would you be able to describe something that is completely tied
to a specific HW this way. E.g., let's assume you'd want to record
some messages that pass between 2 internal processors that match
some criteria?
While you can probably standardize some format from such a matching
[although that too would be very tied to different HW capabilities],
the HW pipeline itself can't be abstracted in this case.
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 16:24 ` Mintz, Yuval
@ 2017-04-20 16:58 ` David Miller
2017-04-20 20:12 ` Mintz, Yuval
2017-04-21 8:11 ` Or Gerlitz
0 siblings, 2 replies; 31+ messages in thread
From: David Miller @ 2017-04-20 16:58 UTC (permalink / raw)
To: Yuval.Mintz
Cc: jiri, gerlitz.or, jeffrey.t.kirsher, mitch.a.williams, netdev,
nhorman, sassmann, jogreene, Chad.Dupuis, Manish.Rangankar
From: "Mintz, Yuval" <Yuval.Mintz@cavium.com>
Date: Thu, 20 Apr 2017 16:24:51 +0000
> the HW pipeline itself can't be abstracted in this case.
I've heard that argument before, and I'm glad Jiri didn't drink the
koolaide and instead wrote dpipe.
^ permalink raw reply [flat|nested] 31+ messages in thread
* RE: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 16:58 ` David Miller
@ 2017-04-20 20:12 ` Mintz, Yuval
2017-04-20 20:35 ` David Miller
2017-04-21 8:11 ` Or Gerlitz
1 sibling, 1 reply; 31+ messages in thread
From: Mintz, Yuval @ 2017-04-20 20:12 UTC (permalink / raw)
To: David Miller
Cc: jiri@resnulli.us, gerlitz.or@gmail.com,
jeffrey.t.kirsher@intel.com, mitch.a.williams@intel.com,
netdev@vger.kernel.org, nhorman@redhat.com, sassmann@redhat.com,
jogreene@redhat.com, Dupuis, Chad, Rangankar, Manish
> > the HW pipeline itself can't be abstracted in this case.
>
> I've heard that argument before, and I'm glad Jiri didn't drink the koolaide
> and instead wrote dpipe.
Perhaps "can't" was the wrong term to use, but I still believe there's an
inherent problem with applying the dpipe approach here.
dpipe abstraction is meant [And I admit that I'm not that familiar with its
fine-grain details, so feel free to correct me] to give an abstract
representation of functional things, I.e., bits in the HW pipeline the user
might care about and can gain something from sharing information
about it - decision-based statistics [such as in Jiri's initial submission] being
an immediate candidate for benefiting from such.
For the case of debugging you *could* do the same, but at a huge
implementation and code-bloat costs -
Let's assume user doesn't use 'magic values' for debug gathering,
Instead having some HW abstraction it can manipulate intelligibly.
What it means is that level of configurability is limited by the complexity
of the abstraction, and you have to recall the vast majority of the
abstraction would be in-depth about HW-specific bridges and signals
the user cares nothing about [with the exception of debug-data collection].
I surely wouldn't want to write a million lines of code just to provide such
a detailed abstraction.
.
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 20:12 ` Mintz, Yuval
@ 2017-04-20 20:35 ` David Miller
2017-04-20 20:50 ` Mintz, Yuval
0 siblings, 1 reply; 31+ messages in thread
From: David Miller @ 2017-04-20 20:35 UTC (permalink / raw)
To: Yuval.Mintz
Cc: jiri, gerlitz.or, jeffrey.t.kirsher, mitch.a.williams, netdev,
nhorman, sassmann, jogreene, Chad.Dupuis, Manish.Rangankar
From: "Mintz, Yuval" <Yuval.Mintz@cavium.com>
Date: Thu, 20 Apr 2017 20:12:59 +0000
> I surely wouldn't want to write a million lines of code just to
> provide such a detailed abstraction.
So now the argument has gone from "it's not possible" to "it's going
to be a HUGE amount of code."
I wonder what the next argument will be when I shoot this one down.
If you simply don't want to do the hard work necessary to implement
this properly, which is fine, just say so.
^ permalink raw reply [flat|nested] 31+ messages in thread
* RE: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 20:35 ` David Miller
@ 2017-04-20 20:50 ` Mintz, Yuval
2017-04-20 20:54 ` David Miller
0 siblings, 1 reply; 31+ messages in thread
From: Mintz, Yuval @ 2017-04-20 20:50 UTC (permalink / raw)
To: David Miller
Cc: jiri@resnulli.us, gerlitz.or@gmail.com,
jeffrey.t.kirsher@intel.com, mitch.a.williams@intel.com,
netdev@vger.kernel.org, nhorman@redhat.com, sassmann@redhat.com,
jogreene@redhat.com, Dupuis, Chad, Rangankar, Manish
> > I surely wouldn't want to write a million lines of code just to
> > provide such a detailed abstraction.
>
> So now the argument has gone from "it's not possible" to "it's going to be a
> HUGE amount of code."
Nothing here's is strictly impossible - you can always create a simulated HW
model in SW, and have a one-to-one mapping with the HW entities.
But "unreasonable amount of code" comes very close to it.
> I wonder what the next argument will be when I shoot this one down.
>
> If you simply don't want to do the hard work necessary to implement this
> properly, which is fine, just say so.
Nope, I don't mind doing the work. What I would appreciate is some
constructive suggestions on a seemingly feasible approach that might work.
I feel like I've been bashed enough for one day. :-)
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 20:50 ` Mintz, Yuval
@ 2017-04-20 20:54 ` David Miller
0 siblings, 0 replies; 31+ messages in thread
From: David Miller @ 2017-04-20 20:54 UTC (permalink / raw)
To: Yuval.Mintz
Cc: jiri, gerlitz.or, jeffrey.t.kirsher, mitch.a.williams, netdev,
nhorman, sassmann, jogreene, Chad.Dupuis, Manish.Rangankar
From: "Mintz, Yuval" <Yuval.Mintz@cavium.com>
Date: Thu, 20 Apr 2017 20:50:55 +0000
> Nope, I don't mind doing the work. What I would appreciate is some
> constructive suggestions on a seemingly feasible approach that might work.
> I feel like I've been bashed enough for one day. :-)
Well, it seems like the thing that you want to do is pull items out of
an object. Start with that as an abstraction and see where it goes.
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 16:58 ` David Miller
2017-04-20 20:12 ` Mintz, Yuval
@ 2017-04-21 8:11 ` Or Gerlitz
1 sibling, 0 replies; 31+ messages in thread
From: Or Gerlitz @ 2017-04-21 8:11 UTC (permalink / raw)
To: David Miller
Cc: Yuval.Mintz, Jiri Pirko, Jeff Kirsher, Mitch Williams,
Linux Netdev List, nhorman@redhat.com, sassmann@redhat.com,
jogreene@redhat.com, Chad.Dupuis, Manish Rangankar
On Thu, Apr 20, 2017 at 7:58 PM, David Miller <davem@davemloft.net> wrote:
> From: "Mintz, Yuval" <Yuval.Mintz@cavium.com>
> Date: Thu, 20 Apr 2017 16:24:51 +0000
>
>> the HW pipeline itself can't be abstracted in this case.
>
> I've heard that argument before, and I'm glad Jiri didn't drink the
> koolaide and instead wrote dpipe.
donno re drinking but re writing s/jiri/arkadi/
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 10:09 ` Mintz, Yuval
2017-04-20 11:39 ` Jiri Pirko
@ 2017-04-20 15:20 ` David Miller
2017-04-20 22:25 ` Williams, Mitch A
1 sibling, 1 reply; 31+ messages in thread
From: David Miller @ 2017-04-20 15:20 UTC (permalink / raw)
To: Yuval.Mintz
Cc: gerlitz.or, jeffrey.t.kirsher, mitch.a.williams, netdev, nhorman,
sassmann, jogreene
From: "Mintz, Yuval" <Yuval.Mintz@cavium.com>
Date: Thu, 20 Apr 2017 10:09:28 +0000
>> > Dump some internal state about VFs through debugfs. This provides
>> > information not available with 'ip link show'.
>>
>> such as?
>>
>> donnwantobethedebugfspolice, but still, in the 2-3 times we tried to push
>> debugfs to MLNX NIC drivers, Dave disallowed that, and lately the switch
>> team even went further and deleted that portion of the mlxsw driver -- all to
>> all, I don't see much point for these type of changes, thoughts?
>
> Don't want to hikjack your thread, but continuing this topic -
> Is there some flat-out disapproval for debugfs in net-next now?
>
> We're currently internally engaged with adding qed support for register dumps
> [~equivalents for `ethtool -d' outputs] through debugfs, on behalf of storage
> drivers [qedi/qedf] lacking the API for doing that.
I really hate to see debugfs things in networking drivers. It's a
complete cop out for doing things properly.
I push back, but I can only fight too much. If people want to keep
adding stupid poorly designed crap endlessly to their drivers there
is only so much I can do...
^ permalink raw reply [flat|nested] 31+ messages in thread
* RE: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 15:20 ` David Miller
@ 2017-04-20 22:25 ` Williams, Mitch A
2017-04-21 18:12 ` David Miller
0 siblings, 1 reply; 31+ messages in thread
From: Williams, Mitch A @ 2017-04-20 22:25 UTC (permalink / raw)
To: 'David Miller', Yuval.Mintz@cavium.com
Cc: gerlitz.or@gmail.com, Kirsher, Jeffrey T, netdev@vger.kernel.org,
nhorman@redhat.com, sassmann@redhat.com, jogreene@redhat.com
> -----Original Message-----
> From: netdev-owner@vger.kernel.org [mailto:netdev-owner@vger.kernel.org] On
> Behalf Of David Miller
> Sent: Thursday, April 20, 2017 8:20 AM
> To: Yuval.Mintz@cavium.com
> Cc: gerlitz.or@gmail.com; Kirsher, Jeffrey T <jeffrey.t.kirsher@intel.com>;
> Williams, Mitch A <mitch.a.williams@intel.com>; netdev@vger.kernel.org;
> nhorman@redhat.com; sassmann@redhat.com; jogreene@redhat.com
> Subject: Re: [net-next 04/14] i40e: dump VF information in debugfs
>
> From: "Mintz, Yuval" <Yuval.Mintz@cavium.com>
> Date: Thu, 20 Apr 2017 10:09:28 +0000
>
> >> > Dump some internal state about VFs through debugfs. This provides
> >> > information not available with 'ip link show'.
> >>
> >> such as?
> >>
> >> donnwantobethedebugfspolice, but still, in the 2-3 times we tried to push
> >> debugfs to MLNX NIC drivers, Dave disallowed that, and lately the switch
> >> team even went further and deleted that portion of the mlxsw driver --
> all to
> >> all, I don't see much point for these type of changes, thoughts?
> >
> > Don't want to hikjack your thread, but continuing this topic -
> > Is there some flat-out disapproval for debugfs in net-next now?
> >
> > We're currently internally engaged with adding qed support for register
> dumps
> > [~equivalents for `ethtool -d' outputs] through debugfs, on behalf of
> storage
> > drivers [qedi/qedf] lacking the API for doing that.
>
> I really hate to see debugfs things in networking drivers. It's a
> complete cop out for doing things properly.
>
> I push back, but I can only fight too much. If people want to keep
> adding stupid poorly designed crap endlessly to their drivers there
> is only so much I can do...
Honestly, Dave, I try very hard not to put random crap into debugfs. I felt that there was enough utility here to send this one up.
We already have debugfs hooks upstream that allow us to dump information about a specific VSI. This just adds a tiny bit to that to allow us to link a VF with its specific VSI. It's the missing link, if you will.
I'm not adding backdoor hooks to do naughty things, I'm just (very slightly) enhancing what's already there.
This particular piece has been very useful for my debugging, and we have customers that have asked for it as well.
If you still want to reject this, I won't fight and cry and stomp my little feet. But I honestly think that this a) has genuine utility, b) is fairly device specific, and c) isn't crap.
-Mitch
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [net-next 04/14] i40e: dump VF information in debugfs
2017-04-20 22:25 ` Williams, Mitch A
@ 2017-04-21 18:12 ` David Miller
0 siblings, 0 replies; 31+ messages in thread
From: David Miller @ 2017-04-21 18:12 UTC (permalink / raw)
To: mitch.a.williams
Cc: Yuval.Mintz, gerlitz.or, jeffrey.t.kirsher, netdev, nhorman,
sassmann, jogreene
From: "Williams, Mitch A" <mitch.a.williams@intel.com>
Date: Thu, 20 Apr 2017 22:25:03 +0000
> I'm not adding backdoor hooks to do naughty things, I'm just (very
> slightly) enhancing what's already there.
I'm not going to drop this series because of this.
I'm just sounding the warning that I probably have to start pushing
back harder in the future because seriously.. debugfs is not deployed
production and in fact a lot of people disable it for security
reasons.
Even tracing got out of debugfs for this reason.
^ permalink raw reply [flat|nested] 31+ messages in thread
* [net-next 05/14] i40e/i40evf: Add tracepoints
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (3 preceding siblings ...)
2017-04-20 1:57 ` [net-next 04/14] i40e: dump VF information in debugfs Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 06/14] i40e: new AQ commands Jeff Kirsher
` (9 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Scott Peterson, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Scott Peterson <scott.d.peterson@intel.com>
This patch adds tracepoints to the i40e and i40evf drivers to which
BPF programs can be attached for feature testing and verification.
It's expected that an attached BPF program will identify and count or
log some interesting subset of traffic. The bcc-tools package is
helpful there for containing all the BPF arcana in a handy Python
wrapper. Though you can make these tracepoints log trace messages, the
messages themselves probably won't be very useful (other to verify the
tracepoint is being called while you're debugging your BPF program).
The idea here is that tracepoints have such low performance cost when
disabled that we can leave these in the upstream drivers. This may
eventually enable the instrumentation of unmodified customer systems
should the need arise to verify a NIC feature is working as expected.
In general this enables one set of feature verification tools to be
used on these drivers whether they're built with the kernel or
separately.
Users are advised against using these tracepoints for anything other
than a diagnostic tool. They have a performance impact when enabled,
and their exact placement and form may change as we see how well they
work in practice for the purposes above.
Change-ID: Id6014a7322c0e6d08068114dd20bd156f2f6435e
Signed-off-by: Scott Peterson <scott.d.peterson@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/Makefile | 3 +
drivers/net/ethernet/intel/i40e/i40e_main.c | 6 +
drivers/net/ethernet/intel/i40e/i40e_trace.h | 229 ++++++++++++++++++++++++
drivers/net/ethernet/intel/i40e/i40e_txrx.c | 9 +
drivers/net/ethernet/intel/i40evf/Makefile | 3 +
drivers/net/ethernet/intel/i40evf/i40e_trace.h | 229 ++++++++++++++++++++++++
drivers/net/ethernet/intel/i40evf/i40e_txrx.c | 9 +
drivers/net/ethernet/intel/i40evf/i40evf_main.c | 7 +
8 files changed, 495 insertions(+)
create mode 100644 drivers/net/ethernet/intel/i40e/i40e_trace.h
create mode 100644 drivers/net/ethernet/intel/i40evf/i40e_trace.h
diff --git a/drivers/net/ethernet/intel/i40e/Makefile b/drivers/net/ethernet/intel/i40e/Makefile
index 4f454d364d0d..3da482c3d68d 100644
--- a/drivers/net/ethernet/intel/i40e/Makefile
+++ b/drivers/net/ethernet/intel/i40e/Makefile
@@ -28,6 +28,9 @@
# Makefile for the Intel(R) Ethernet Connection XL710 (i40e.ko) driver
#
+ccflags-y += -I$(src)
+subdir-ccflags-y += -I$(src)
+
obj-$(CONFIG_I40E) += i40e.o
i40e-objs := i40e_main.o \
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index b6ec9beeebff..5e625e0c73ac 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -32,6 +32,12 @@
#include "i40e.h"
#include "i40e_diag.h"
#include <net/udp_tunnel.h>
+/* All i40e tracepoints are defined by the include below, which
+ * must be included exactly once across the whole kernel with
+ * CREATE_TRACE_POINTS defined
+ */
+#define CREATE_TRACE_POINTS
+#include "i40e_trace.h"
const char i40e_driver_name[] = "i40e";
static const char i40e_driver_string[] =
diff --git a/drivers/net/ethernet/intel/i40e/i40e_trace.h b/drivers/net/ethernet/intel/i40e/i40e_trace.h
new file mode 100644
index 000000000000..d3e55f54a05e
--- /dev/null
+++ b/drivers/net/ethernet/intel/i40e/i40e_trace.h
@@ -0,0 +1,229 @@
+/*******************************************************************************
+ *
+ * Intel(R) 40-10 Gigabit Ethernet Connection Network Driver
+ * Copyright(c) 2013 - 2017 Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
+ * more details.
+ *
+ * The full GNU General Public License is included in this distribution in
+ * the file called "COPYING".
+ *
+ * Contact Information:
+ * e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+ * Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+ *
+ ******************************************************************************/
+
+/* Modeled on trace-events-sample.h */
+
+/* The trace subsystem name for i40e will be "i40e".
+ *
+ * This file is named i40e_trace.h.
+ *
+ * Since this include file's name is different from the trace
+ * subsystem name, we'll have to define TRACE_INCLUDE_FILE at the end
+ * of this file.
+ */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM i40e
+
+/* See trace-events-sample.h for a detailed description of why this
+ * guard clause is different from most normal include files.
+ */
+#if !defined(_I40E_TRACE_H_) || defined(TRACE_HEADER_MULTI_READ)
+#define _I40E_TRACE_H_
+
+#include <linux/tracepoint.h>
+
+/**
+ * i40e_trace() macro enables shared code to refer to trace points
+ * like:
+ *
+ * trace_i40e{,vf}_example(args...)
+ *
+ * ... as:
+ *
+ * i40e_trace(example, args...)
+ *
+ * ... to resolve to the PF or VF version of the tracepoint without
+ * ifdefs, and to allow tracepoints to be disabled entirely at build
+ * time.
+ *
+ * Trace point should always be referred to in the driver via this
+ * macro.
+ *
+ * Similarly, i40e_trace_enabled(trace_name) wraps references to
+ * trace_i40e{,vf}_<trace_name>_enabled() functions.
+ */
+#define _I40E_TRACE_NAME(trace_name) (trace_ ## i40e ## _ ## trace_name)
+#define I40E_TRACE_NAME(trace_name) _I40E_TRACE_NAME(trace_name)
+
+#define i40e_trace(trace_name, args...) I40E_TRACE_NAME(trace_name)(args)
+
+#define i40e_trace_enabled(trace_name) I40E_TRACE_NAME(trace_name##_enabled)()
+
+/* Events common to PF and VF. Corresponding versions will be defined
+ * for both, named trace_i40e_* and trace_i40evf_*. The i40e_trace()
+ * macro above will select the right trace point name for the driver
+ * being built from shared code.
+ */
+
+/* Events related to a vsi & ring */
+DECLARE_EVENT_CLASS(
+ i40e_tx_template,
+
+ TP_PROTO(struct i40e_ring *ring,
+ struct i40e_tx_desc *desc,
+ struct i40e_tx_buffer *buf),
+
+ TP_ARGS(ring, desc, buf),
+
+ /* The convention here is to make the first fields in the
+ * TP_STRUCT match the TP_PROTO exactly. This enables the use
+ * of the args struct generated by the tplist tool (from the
+ * bcc-tools package) to be used for those fields. To access
+ * fields other than the tracepoint args will require the
+ * tplist output to be adjusted.
+ */
+ TP_STRUCT__entry(
+ __field(void*, ring)
+ __field(void*, desc)
+ __field(void*, buf)
+ __string(devname, ring->netdev->name)
+ ),
+
+ TP_fast_assign(
+ __entry->ring = ring;
+ __entry->desc = desc;
+ __entry->buf = buf;
+ __assign_str(devname, ring->netdev->name);
+ ),
+
+ TP_printk(
+ "netdev: %s ring: %p desc: %p buf %p",
+ __get_str(devname), __entry->ring,
+ __entry->desc, __entry->buf)
+);
+
+DEFINE_EVENT(
+ i40e_tx_template, i40e_clean_tx_irq,
+ TP_PROTO(struct i40e_ring *ring,
+ struct i40e_tx_desc *desc,
+ struct i40e_tx_buffer *buf),
+
+ TP_ARGS(ring, desc, buf));
+
+DEFINE_EVENT(
+ i40e_tx_template, i40e_clean_tx_irq_unmap,
+ TP_PROTO(struct i40e_ring *ring,
+ struct i40e_tx_desc *desc,
+ struct i40e_tx_buffer *buf),
+
+ TP_ARGS(ring, desc, buf));
+
+DECLARE_EVENT_CLASS(
+ i40e_rx_template,
+
+ TP_PROTO(struct i40e_ring *ring,
+ union i40e_32byte_rx_desc *desc,
+ struct sk_buff *skb),
+
+ TP_ARGS(ring, desc, skb),
+
+ TP_STRUCT__entry(
+ __field(void*, ring)
+ __field(void*, desc)
+ __field(void*, skb)
+ __string(devname, ring->netdev->name)
+ ),
+
+ TP_fast_assign(
+ __entry->ring = ring;
+ __entry->desc = desc;
+ __entry->skb = skb;
+ __assign_str(devname, ring->netdev->name);
+ ),
+
+ TP_printk(
+ "netdev: %s ring: %p desc: %p skb %p",
+ __get_str(devname), __entry->ring,
+ __entry->desc, __entry->skb)
+);
+
+DEFINE_EVENT(
+ i40e_rx_template, i40e_clean_rx_irq,
+ TP_PROTO(struct i40e_ring *ring,
+ union i40e_32byte_rx_desc *desc,
+ struct sk_buff *skb),
+
+ TP_ARGS(ring, desc, skb));
+
+DEFINE_EVENT(
+ i40e_rx_template, i40e_clean_rx_irq_rx,
+ TP_PROTO(struct i40e_ring *ring,
+ union i40e_32byte_rx_desc *desc,
+ struct sk_buff *skb),
+
+ TP_ARGS(ring, desc, skb));
+
+DECLARE_EVENT_CLASS(
+ i40e_xmit_template,
+
+ TP_PROTO(struct sk_buff *skb,
+ struct i40e_ring *ring),
+
+ TP_ARGS(skb, ring),
+
+ TP_STRUCT__entry(
+ __field(void*, skb)
+ __field(void*, ring)
+ __string(devname, ring->netdev->name)
+ ),
+
+ TP_fast_assign(
+ __entry->skb = skb;
+ __entry->ring = ring;
+ __assign_str(devname, ring->netdev->name);
+ ),
+
+ TP_printk(
+ "netdev: %s skb: %p ring: %p",
+ __get_str(devname), __entry->skb,
+ __entry->ring)
+);
+
+DEFINE_EVENT(
+ i40e_xmit_template, i40e_xmit_frame_ring,
+ TP_PROTO(struct sk_buff *skb,
+ struct i40e_ring *ring),
+
+ TP_ARGS(skb, ring));
+
+DEFINE_EVENT(
+ i40e_xmit_template, i40e_xmit_frame_ring_drop,
+ TP_PROTO(struct sk_buff *skb,
+ struct i40e_ring *ring),
+
+ TP_ARGS(skb, ring));
+
+/* Events unique to the PF. */
+
+#endif /* _I40E_TRACE_H_ */
+/* This must be outside ifdef _I40E_TRACE_H */
+
+/* This trace include file is not located in the .../include/trace
+ * with the kernel tracepoint definitions, because we're a loadable
+ * module.
+ */
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH .
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_FILE i40e_trace
+#include <trace/define_trace.h>
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index 843d7ffe1784..1531a0f9fcc6 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -27,6 +27,7 @@
#include <linux/prefetch.h>
#include <net/busy_poll.h>
#include "i40e.h"
+#include "i40e_trace.h"
#include "i40e_prototype.h"
static inline __le64 build_ctob(u32 td_cmd, u32 td_offset, unsigned int size,
@@ -765,6 +766,7 @@ static bool i40e_clean_tx_irq(struct i40e_vsi *vsi,
/* prevent any other reads prior to eop_desc */
read_barrier_depends();
+ i40e_trace(clean_tx_irq, tx_ring, tx_desc, tx_buf);
/* we have caught up to head, no work left to do */
if (tx_head == tx_desc)
break;
@@ -791,6 +793,8 @@ static bool i40e_clean_tx_irq(struct i40e_vsi *vsi,
/* unmap remaining buffers */
while (tx_desc != eop_desc) {
+ i40e_trace(clean_tx_irq_unmap,
+ tx_ring, tx_desc, tx_buf);
tx_buf++;
tx_desc++;
@@ -2005,6 +2009,7 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
if (!size)
break;
+ i40e_trace(clean_rx_irq, rx_ring, rx_desc, skb);
rx_buffer = i40e_get_rx_buffer(rx_ring, size);
/* retrieve a buffer from the ring */
@@ -2057,6 +2062,7 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
vlan_tag = (qword & BIT(I40E_RX_DESC_STATUS_L2TAG1P_SHIFT)) ?
le16_to_cpu(rx_desc->wb.qword0.lo_dword.l2tag1) : 0;
+ i40e_trace(clean_rx_irq_rx, rx_ring, rx_desc, skb);
i40e_receive_skb(rx_ring, skb, vlan_tag);
skb = NULL;
@@ -3138,6 +3144,8 @@ static netdev_tx_t i40e_xmit_frame_ring(struct sk_buff *skb,
/* prefetch the data, we'll need it later */
prefetch(skb->data);
+ i40e_trace(xmit_frame_ring, skb, tx_ring);
+
count = i40e_xmit_descriptor_count(skb);
if (i40e_chk_linearize(skb, count)) {
if (__skb_linearize(skb)) {
@@ -3216,6 +3224,7 @@ static netdev_tx_t i40e_xmit_frame_ring(struct sk_buff *skb,
return NETDEV_TX_OK;
out_drop:
+ i40e_trace(xmit_frame_ring_drop, first->skb, tx_ring);
dev_kfree_skb_any(first->skb);
first->skb = NULL;
return NETDEV_TX_OK;
diff --git a/drivers/net/ethernet/intel/i40evf/Makefile b/drivers/net/ethernet/intel/i40evf/Makefile
index 827c7a6ed0ba..a393f4a07f06 100644
--- a/drivers/net/ethernet/intel/i40evf/Makefile
+++ b/drivers/net/ethernet/intel/i40evf/Makefile
@@ -29,6 +29,9 @@
#
#
+ccflags-y += -I$(src)
+subdir-ccflags-y += -I$(src)
+
obj-$(CONFIG_I40EVF) += i40evf.o
i40evf-objs := i40evf_main.o i40evf_ethtool.o i40evf_virtchnl.o \
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_trace.h b/drivers/net/ethernet/intel/i40evf/i40e_trace.h
new file mode 100644
index 000000000000..9a5100b2b7c7
--- /dev/null
+++ b/drivers/net/ethernet/intel/i40evf/i40e_trace.h
@@ -0,0 +1,229 @@
+/*******************************************************************************
+ *
+ * Intel(R) 40-10 Gigabit Ethernet Virtual Function Driver
+ * Copyright(c) 2013 - 2017 Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
+ * more details.
+ *
+ * The full GNU General Public License is included in this distribution in
+ * the file called "COPYING".
+ *
+ * Contact Information:
+ * e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+ * Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+ *
+ ******************************************************************************/
+
+/* Modeled on trace-events-sample.h */
+
+/* The trace subsystem name for i40evf will be "i40evf".
+ *
+ * This file is named i40e_trace.h.
+ *
+ * Since this include file's name is different from the trace
+ * subsystem name, we'll have to define TRACE_INCLUDE_FILE at the end
+ * of this file.
+ */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM i40evf
+
+/* See trace-events-sample.h for a detailed description of why this
+ * guard clause is different from most normal include files.
+ */
+#if !defined(_I40E_TRACE_H_) || defined(TRACE_HEADER_MULTI_READ)
+#define _I40E_TRACE_H_
+
+#include <linux/tracepoint.h>
+
+/**
+ * i40e_trace() macro enables shared code to refer to trace points
+ * like:
+ *
+ * trace_i40e{,vf}_example(args...)
+ *
+ * ... as:
+ *
+ * i40e_trace(example, args...)
+ *
+ * ... to resolve to the PF or VF version of the tracepoint without
+ * ifdefs, and to allow tracepoints to be disabled entirely at build
+ * time.
+ *
+ * Trace point should always be referred to in the driver via this
+ * macro.
+ *
+ * Similarly, i40e_trace_enabled(trace_name) wraps references to
+ * trace_i40e{,vf}_<trace_name>_enabled() functions.
+ */
+#define _I40E_TRACE_NAME(trace_name) (trace_ ## i40evf ## _ ## trace_name)
+#define I40E_TRACE_NAME(trace_name) _I40E_TRACE_NAME(trace_name)
+
+#define i40e_trace(trace_name, args...) I40E_TRACE_NAME(trace_name)(args)
+
+#define i40e_trace_enabled(trace_name) I40E_TRACE_NAME(trace_name##_enabled)()
+
+/* Events common to PF and VF. Corresponding versions will be defined
+ * for both, named trace_i40e_* and trace_i40evf_*. The i40e_trace()
+ * macro above will select the right trace point name for the driver
+ * being built from shared code.
+ */
+
+/* Events related to a vsi & ring */
+DECLARE_EVENT_CLASS(
+ i40evf_tx_template,
+
+ TP_PROTO(struct i40e_ring *ring,
+ struct i40e_tx_desc *desc,
+ struct i40e_tx_buffer *buf),
+
+ TP_ARGS(ring, desc, buf),
+
+ /* The convention here is to make the first fields in the
+ * TP_STRUCT match the TP_PROTO exactly. This enables the use
+ * of the args struct generated by the tplist tool (from the
+ * bcc-tools package) to be used for those fields. To access
+ * fields other than the tracepoint args will require the
+ * tplist output to be adjusted.
+ */
+ TP_STRUCT__entry(
+ __field(void*, ring)
+ __field(void*, desc)
+ __field(void*, buf)
+ __string(devname, ring->netdev->name)
+ ),
+
+ TP_fast_assign(
+ __entry->ring = ring;
+ __entry->desc = desc;
+ __entry->buf = buf;
+ __assign_str(devname, ring->netdev->name);
+ ),
+
+ TP_printk(
+ "netdev: %s ring: %p desc: %p buf %p",
+ __get_str(devname), __entry->ring,
+ __entry->desc, __entry->buf)
+);
+
+DEFINE_EVENT(
+ i40evf_tx_template, i40evf_clean_tx_irq,
+ TP_PROTO(struct i40e_ring *ring,
+ struct i40e_tx_desc *desc,
+ struct i40e_tx_buffer *buf),
+
+ TP_ARGS(ring, desc, buf));
+
+DEFINE_EVENT(
+ i40evf_tx_template, i40evf_clean_tx_irq_unmap,
+ TP_PROTO(struct i40e_ring *ring,
+ struct i40e_tx_desc *desc,
+ struct i40e_tx_buffer *buf),
+
+ TP_ARGS(ring, desc, buf));
+
+DECLARE_EVENT_CLASS(
+ i40evf_rx_template,
+
+ TP_PROTO(struct i40e_ring *ring,
+ union i40e_32byte_rx_desc *desc,
+ struct sk_buff *skb),
+
+ TP_ARGS(ring, desc, skb),
+
+ TP_STRUCT__entry(
+ __field(void*, ring)
+ __field(void*, desc)
+ __field(void*, skb)
+ __string(devname, ring->netdev->name)
+ ),
+
+ TP_fast_assign(
+ __entry->ring = ring;
+ __entry->desc = desc;
+ __entry->skb = skb;
+ __assign_str(devname, ring->netdev->name);
+ ),
+
+ TP_printk(
+ "netdev: %s ring: %p desc: %p skb %p",
+ __get_str(devname), __entry->ring,
+ __entry->desc, __entry->skb)
+);
+
+DEFINE_EVENT(
+ i40evf_rx_template, i40evf_clean_rx_irq,
+ TP_PROTO(struct i40e_ring *ring,
+ union i40e_32byte_rx_desc *desc,
+ struct sk_buff *skb),
+
+ TP_ARGS(ring, desc, skb));
+
+DEFINE_EVENT(
+ i40evf_rx_template, i40evf_clean_rx_irq_rx,
+ TP_PROTO(struct i40e_ring *ring,
+ union i40e_32byte_rx_desc *desc,
+ struct sk_buff *skb),
+
+ TP_ARGS(ring, desc, skb));
+
+DECLARE_EVENT_CLASS(
+ i40evf_xmit_template,
+
+ TP_PROTO(struct sk_buff *skb,
+ struct i40e_ring *ring),
+
+ TP_ARGS(skb, ring),
+
+ TP_STRUCT__entry(
+ __field(void*, skb)
+ __field(void*, ring)
+ __string(devname, ring->netdev->name)
+ ),
+
+ TP_fast_assign(
+ __entry->skb = skb;
+ __entry->ring = ring;
+ __assign_str(devname, ring->netdev->name);
+ ),
+
+ TP_printk(
+ "netdev: %s skb: %p ring: %p",
+ __get_str(devname), __entry->skb,
+ __entry->ring)
+);
+
+DEFINE_EVENT(
+ i40evf_xmit_template, i40evf_xmit_frame_ring,
+ TP_PROTO(struct sk_buff *skb,
+ struct i40e_ring *ring),
+
+ TP_ARGS(skb, ring));
+
+DEFINE_EVENT(
+ i40evf_xmit_template, i40evf_xmit_frame_ring_drop,
+ TP_PROTO(struct sk_buff *skb,
+ struct i40e_ring *ring),
+
+ TP_ARGS(skb, ring));
+
+/* Events unique to the VF. */
+
+#endif /* _I40E_TRACE_H_ */
+/* This must be outside ifdef _I40E_TRACE_H */
+
+/* This trace include file is not located in the .../include/trace
+ * with the kernel tracepoint definitions, because we're a loadable
+ * module.
+ */
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH .
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_FILE i40e_trace
+#include <trace/define_trace.h>
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
index 80931e3f9445..34e96d98251a 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
@@ -28,6 +28,7 @@
#include <net/busy_poll.h>
#include "i40evf.h"
+#include "i40e_trace.h"
#include "i40e_prototype.h"
static inline __le64 build_ctob(u32 td_cmd, u32 td_offset, unsigned int size,
@@ -180,6 +181,7 @@ static bool i40e_clean_tx_irq(struct i40e_vsi *vsi,
/* prevent any other reads prior to eop_desc */
read_barrier_depends();
+ i40e_trace(clean_tx_irq, tx_ring, tx_desc, tx_buf);
/* if the descriptor isn't done, no work yet to do */
if (!(eop_desc->cmd_type_offset_bsz &
cpu_to_le64(I40E_TX_DESC_DTYPE_DESC_DONE)))
@@ -207,6 +209,8 @@ static bool i40e_clean_tx_irq(struct i40e_vsi *vsi,
/* unmap remaining buffers */
while (tx_desc != eop_desc) {
+ i40e_trace(clean_tx_irq_unmap,
+ tx_ring, tx_desc, tx_buf);
tx_buf++;
tx_desc++;
@@ -1329,6 +1333,7 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
if (!size)
break;
+ i40e_trace(clean_rx_irq, rx_ring, rx_desc, skb);
rx_buffer = i40e_get_rx_buffer(rx_ring, size);
/* retrieve a buffer from the ring */
@@ -1382,6 +1387,7 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
vlan_tag = (qword & BIT(I40E_RX_DESC_STATUS_L2TAG1P_SHIFT)) ?
le16_to_cpu(rx_desc->wb.qword0.lo_dword.l2tag1) : 0;
+ i40e_trace(clean_rx_irq_rx, rx_ring, rx_desc, skb);
i40e_receive_skb(rx_ring, skb, vlan_tag);
skb = NULL;
@@ -2223,6 +2229,8 @@ static netdev_tx_t i40e_xmit_frame_ring(struct sk_buff *skb,
/* prefetch the data, we'll need it later */
prefetch(skb->data);
+ i40e_trace(xmit_frame_ring, skb, tx_ring);
+
count = i40e_xmit_descriptor_count(skb);
if (i40e_chk_linearize(skb, count)) {
if (__skb_linearize(skb)) {
@@ -2290,6 +2298,7 @@ static netdev_tx_t i40e_xmit_frame_ring(struct sk_buff *skb,
return NETDEV_TX_OK;
out_drop:
+ i40e_trace(xmit_frame_ring_drop, first->skb, tx_ring);
dev_kfree_skb_any(first->skb);
first->skb = NULL;
return NETDEV_TX_OK;
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_main.c b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
index 671913c74bf8..5915273c372f 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c
+++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
@@ -27,6 +27,13 @@
#include "i40evf.h"
#include "i40e_prototype.h"
#include "i40evf_client.h"
+/* All i40evf tracepoints are defined by the include below, which must
+ * be included exactly once across the whole kernel with
+ * CREATE_TRACE_POINTS defined
+ */
+#define CREATE_TRACE_POINTS
+#include "i40e_trace.h"
+
static int i40evf_setup_all_tx_resources(struct i40evf_adapter *adapter);
static int i40evf_setup_all_rx_resources(struct i40evf_adapter *adapter);
static int i40evf_close(struct net_device *netdev);
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 06/14] i40e: new AQ commands
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (4 preceding siblings ...)
2017-04-20 1:57 ` [net-next 05/14] i40e/i40evf: Add tracepoints Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 07/14] i40e: don't hold RTNL lock while waiting for VF reset to finish Jeff Kirsher
` (8 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Jingjing Wu, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Jingjing Wu <jingjing.wu@intel.com>
Add admin queue functions for Pipeline Personalization Profile AQ
commands:
- Write Recipe Command buffer (Opcode: 0x0270)
- Get Applied Profiles list (Opcode: 0x0271)
Change-ID: I558b4145364140f624013af48d4bbf79d21ebb0d
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e_adminq_cmd.h | 34 ++++
drivers/net/ethernet/intel/i40e/i40e_common.c | 212 +++++++++++++++++++++
drivers/net/ethernet/intel/i40e/i40e_prototype.h | 17 ++
drivers/net/ethernet/intel/i40e/i40e_type.h | 80 ++++++++
.../net/ethernet/intel/i40evf/i40e_adminq_cmd.h | 34 ++++
drivers/net/ethernet/intel/i40evf/i40e_common.c | 212 +++++++++++++++++++++
drivers/net/ethernet/intel/i40evf/i40e_prototype.h | 17 ++
drivers/net/ethernet/intel/i40evf/i40e_type.h | 80 ++++++++
8 files changed, 686 insertions(+)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_adminq_cmd.h b/drivers/net/ethernet/intel/i40e/i40e_adminq_cmd.h
index 251074c677c4..5eb04114e13f 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/i40e/i40e_adminq_cmd.h
@@ -190,6 +190,10 @@ enum i40e_admin_queue_opc {
i40e_aqc_opc_add_mirror_rule = 0x0260,
i40e_aqc_opc_delete_mirror_rule = 0x0261,
+ /* Pipeline Personalization Profile */
+ i40e_aqc_opc_write_personalization_profile = 0x0270,
+ i40e_aqc_opc_get_personalization_profile_list = 0x0271,
+
/* DCB commands */
i40e_aqc_opc_dcb_ignore_pfc = 0x0301,
i40e_aqc_opc_dcb_updated = 0x0302,
@@ -1431,6 +1435,36 @@ struct i40e_aqc_add_delete_mirror_rule_completion {
I40E_CHECK_CMD_LENGTH(i40e_aqc_add_delete_mirror_rule_completion);
+/* Pipeline Personalization Profile */
+struct i40e_aqc_write_personalization_profile {
+ u8 flags;
+ u8 reserved[3];
+ __le32 profile_track_id;
+ __le32 addr_high;
+ __le32 addr_low;
+};
+
+I40E_CHECK_CMD_LENGTH(i40e_aqc_write_personalization_profile);
+
+struct i40e_aqc_write_ppp_resp {
+ __le32 error_offset;
+ __le32 error_info;
+ __le32 addr_high;
+ __le32 addr_low;
+};
+
+struct i40e_aqc_get_applied_profiles {
+ u8 flags;
+#define I40E_AQC_GET_PPP_GET_CONF 0x1
+#define I40E_AQC_GET_PPP_GET_RDPU_CONF 0x2
+ u8 rsv[3];
+ __le32 reserved;
+ __le32 addr_high;
+ __le32 addr_low;
+};
+
+I40E_CHECK_CMD_LENGTH(i40e_aqc_get_applied_profiles);
+
/* DCB 0x03xx*/
/* PFC Ignore (direct 0x0301)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_common.c b/drivers/net/ethernet/intel/i40e/i40e_common.c
index f9db95aa3a20..24f020655291 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_common.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_common.c
@@ -5042,3 +5042,215 @@ void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val)
if (status || use_register)
wr32(hw, reg_addr, reg_val);
}
+
+/**
+ * i40e_aq_write_ppp - Write pipeline personalization profile (ppp)
+ * @hw: pointer to the hw struct
+ * @buff: command buffer (size in bytes = buff_size)
+ * @buff_size: buffer size in bytes
+ * @track_id: package tracking id
+ * @error_offset: returns error offset
+ * @error_info: returns error information
+ * @cmd_details: pointer to command details structure or NULL
+ **/
+enum
+i40e_status_code i40e_aq_write_ppp(struct i40e_hw *hw, void *buff,
+ u16 buff_size, u32 track_id,
+ u32 *error_offset, u32 *error_info,
+ struct i40e_asq_cmd_details *cmd_details)
+{
+ struct i40e_aq_desc desc;
+ struct i40e_aqc_write_personalization_profile *cmd =
+ (struct i40e_aqc_write_personalization_profile *)
+ &desc.params.raw;
+ struct i40e_aqc_write_ppp_resp *resp;
+ i40e_status status;
+
+ i40e_fill_default_direct_cmd_desc(&desc,
+ i40e_aqc_opc_write_personalization_profile);
+
+ desc.flags |= cpu_to_le16(I40E_AQ_FLAG_BUF | I40E_AQ_FLAG_RD);
+ if (buff_size > I40E_AQ_LARGE_BUF)
+ desc.flags |= cpu_to_le16((u16)I40E_AQ_FLAG_LB);
+
+ desc.datalen = cpu_to_le16(buff_size);
+
+ cmd->profile_track_id = cpu_to_le32(track_id);
+
+ status = i40e_asq_send_command(hw, &desc, buff, buff_size, cmd_details);
+ if (!status) {
+ resp = (struct i40e_aqc_write_ppp_resp *)&desc.params.raw;
+ if (error_offset)
+ *error_offset = le32_to_cpu(resp->error_offset);
+ if (error_info)
+ *error_info = le32_to_cpu(resp->error_info);
+ }
+
+ return status;
+}
+
+/**
+ * i40e_aq_get_ppp_list - Read pipeline personalization profile (ppp)
+ * @hw: pointer to the hw struct
+ * @buff: command buffer (size in bytes = buff_size)
+ * @buff_size: buffer size in bytes
+ * @cmd_details: pointer to command details structure or NULL
+ **/
+enum
+i40e_status_code i40e_aq_get_ppp_list(struct i40e_hw *hw, void *buff,
+ u16 buff_size, u8 flags,
+ struct i40e_asq_cmd_details *cmd_details)
+{
+ struct i40e_aq_desc desc;
+ struct i40e_aqc_get_applied_profiles *cmd =
+ (struct i40e_aqc_get_applied_profiles *)&desc.params.raw;
+ i40e_status status;
+
+ i40e_fill_default_direct_cmd_desc(&desc,
+ i40e_aqc_opc_get_personalization_profile_list);
+
+ desc.flags |= cpu_to_le16((u16)I40E_AQ_FLAG_BUF);
+ if (buff_size > I40E_AQ_LARGE_BUF)
+ desc.flags |= cpu_to_le16((u16)I40E_AQ_FLAG_LB);
+ desc.datalen = cpu_to_le16(buff_size);
+
+ cmd->flags = flags;
+
+ status = i40e_asq_send_command(hw, &desc, buff, buff_size, cmd_details);
+
+ return status;
+}
+
+/**
+ * i40e_find_segment_in_package
+ * @segment_type: the segment type to search for (i.e., SEGMENT_TYPE_I40E)
+ * @pkg_hdr: pointer to the package header to be searched
+ *
+ * This function searches a package file for a particular segment type. On
+ * success it returns a pointer to the segment header, otherwise it will
+ * return NULL.
+ **/
+struct i40e_generic_seg_header *
+i40e_find_segment_in_package(u32 segment_type,
+ struct i40e_package_header *pkg_hdr)
+{
+ struct i40e_generic_seg_header *segment;
+ u32 i;
+
+ /* Search all package segments for the requested segment type */
+ for (i = 0; i < pkg_hdr->segment_count; i++) {
+ segment =
+ (struct i40e_generic_seg_header *)((u8 *)pkg_hdr +
+ pkg_hdr->segment_offset[i]);
+
+ if (segment->type == segment_type)
+ return segment;
+ }
+
+ return NULL;
+}
+
+/**
+ * i40e_write_profile
+ * @hw: pointer to the hardware structure
+ * @profile: pointer to the profile segment of the package to be downloaded
+ * @track_id: package tracking id
+ *
+ * Handles the download of a complete package.
+ */
+enum i40e_status_code
+i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile,
+ u32 track_id)
+{
+ i40e_status status = 0;
+ struct i40e_section_table *sec_tbl;
+ struct i40e_profile_section_header *sec = NULL;
+ u32 dev_cnt;
+ u32 vendor_dev_id;
+ u32 *nvm;
+ u32 section_size = 0;
+ u32 offset = 0, info = 0;
+ u32 i;
+
+ if (!track_id) {
+ i40e_debug(hw, I40E_DEBUG_PACKAGE, "Track_id can't be 0.");
+ return I40E_NOT_SUPPORTED;
+ }
+
+ dev_cnt = profile->device_table_count;
+
+ for (i = 0; i < dev_cnt; i++) {
+ vendor_dev_id = profile->device_table[i].vendor_dev_id;
+ if ((vendor_dev_id >> 16) == PCI_VENDOR_ID_INTEL)
+ if (hw->device_id == (vendor_dev_id & 0xFFFF))
+ break;
+ }
+ if (i == dev_cnt) {
+ i40e_debug(hw, I40E_DEBUG_PACKAGE, "Device doesn't support PPP");
+ return I40E_ERR_DEVICE_NOT_SUPPORTED;
+ }
+
+ nvm = (u32 *)&profile->device_table[dev_cnt];
+ sec_tbl = (struct i40e_section_table *)&nvm[nvm[0] + 1];
+
+ for (i = 0; i < sec_tbl->section_count; i++) {
+ sec = (struct i40e_profile_section_header *)((u8 *)profile +
+ sec_tbl->section_offset[i]);
+
+ /* Skip 'AQ', 'note' and 'name' sections */
+ if (sec->section.type != SECTION_TYPE_MMIO)
+ continue;
+
+ section_size = sec->section.size +
+ sizeof(struct i40e_profile_section_header);
+
+ /* Write profile */
+ status = i40e_aq_write_ppp(hw, (void *)sec, (u16)section_size,
+ track_id, &offset, &info, NULL);
+ if (status) {
+ i40e_debug(hw, I40E_DEBUG_PACKAGE,
+ "Failed to write profile: offset %d, info %d",
+ offset, info);
+ break;
+ }
+ }
+ return status;
+}
+
+/**
+ * i40e_add_pinfo_to_list
+ * @hw: pointer to the hardware structure
+ * @profile: pointer to the profile segment of the package
+ * @profile_info_sec: buffer for information section
+ * @track_id: package tracking id
+ *
+ * Register a profile to the list of loaded profiles.
+ */
+enum i40e_status_code
+i40e_add_pinfo_to_list(struct i40e_hw *hw,
+ struct i40e_profile_segment *profile,
+ u8 *profile_info_sec, u32 track_id)
+{
+ i40e_status status = 0;
+ struct i40e_profile_section_header *sec = NULL;
+ struct i40e_profile_info *pinfo;
+ u32 offset = 0, info = 0;
+
+ sec = (struct i40e_profile_section_header *)profile_info_sec;
+ sec->tbl_size = 1;
+ sec->data_end = sizeof(struct i40e_profile_section_header) +
+ sizeof(struct i40e_profile_info);
+ sec->section.type = SECTION_TYPE_INFO;
+ sec->section.offset = sizeof(struct i40e_profile_section_header);
+ sec->section.size = sizeof(struct i40e_profile_info);
+ pinfo = (struct i40e_profile_info *)(profile_info_sec +
+ sec->section.offset);
+ pinfo->track_id = track_id;
+ pinfo->version = profile->version;
+ pinfo->op = I40E_PPP_ADD_TRACKID;
+ memcpy(pinfo->name, profile->name, I40E_PPP_NAME_SIZE);
+
+ status = i40e_aq_write_ppp(hw, (void *)sec, sec->data_end,
+ track_id, &offset, &info, NULL);
+ return status;
+}
diff --git a/drivers/net/ethernet/intel/i40e/i40e_prototype.h b/drivers/net/ethernet/intel/i40e/i40e_prototype.h
index dfc5e5901be5..c56d976cf85a 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_prototype.h
+++ b/drivers/net/ethernet/intel/i40e/i40e_prototype.h
@@ -377,4 +377,21 @@ i40e_status i40e_write_phy_register(struct i40e_hw *hw, u8 page, u16 reg,
u8 i40e_get_phy_address(struct i40e_hw *hw, u8 dev_num);
i40e_status i40e_blink_phy_link_led(struct i40e_hw *hw,
u32 time, u32 interval);
+i40e_status i40e_aq_write_ppp(struct i40e_hw *hw, void *buff,
+ u16 buff_size, u32 track_id,
+ u32 *error_offset, u32 *error_info,
+ struct i40e_asq_cmd_details *cmd_details);
+i40e_status i40e_aq_get_ppp_list(struct i40e_hw *hw, void *buff,
+ u16 buff_size, u8 flags,
+ struct i40e_asq_cmd_details *cmd_details);
+struct i40e_generic_seg_header *
+i40e_find_segment_in_package(u32 segment_type,
+ struct i40e_package_header *pkg_header);
+enum i40e_status_code
+i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *i40e_seg,
+ u32 track_id);
+enum i40e_status_code
+i40e_add_pinfo_to_list(struct i40e_hw *hw,
+ struct i40e_profile_segment *profile,
+ u8 *profile_info_sec, u32 track_id);
#endif /* _I40E_PROTOTYPE_H_ */
diff --git a/drivers/net/ethernet/intel/i40e/i40e_type.h b/drivers/net/ethernet/intel/i40e/i40e_type.h
index 9200f2d9c752..3a18ed13edc4 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_type.h
+++ b/drivers/net/ethernet/intel/i40e/i40e_type.h
@@ -78,6 +78,7 @@ enum i40e_debug_mask {
I40E_DEBUG_DCB = 0x00000400,
I40E_DEBUG_DIAG = 0x00000800,
I40E_DEBUG_FD = 0x00001000,
+ I40E_DEBUG_PACKAGE = 0x00002000,
I40E_DEBUG_IWARP = 0x00F00000,
I40E_DEBUG_AQ_MESSAGE = 0x01000000,
I40E_DEBUG_AQ_DESCRIPTOR = 0x02000000,
@@ -1462,4 +1463,83 @@ struct i40e_lldp_variables {
#define I40E_FLEX_56_MASK (0x1ULL << I40E_FLEX_56_SHIFT)
#define I40E_FLEX_57_SHIFT 6
#define I40E_FLEX_57_MASK (0x1ULL << I40E_FLEX_57_SHIFT)
+
+/* Version format for PPP */
+struct i40e_ppp_version {
+ u8 major;
+ u8 minor;
+ u8 update;
+ u8 draft;
+};
+
+#define I40E_PPP_NAME_SIZE 32
+
+/* Package header */
+struct i40e_package_header {
+ struct i40e_ppp_version version;
+ u32 segment_count;
+ u32 segment_offset[1];
+};
+
+/* Generic segment header */
+struct i40e_generic_seg_header {
+#define SEGMENT_TYPE_METADATA 0x00000001
+#define SEGMENT_TYPE_NOTES 0x00000002
+#define SEGMENT_TYPE_I40E 0x00000011
+#define SEGMENT_TYPE_X722 0x00000012
+ u32 type;
+ struct i40e_ppp_version version;
+ u32 size;
+ char name[I40E_PPP_NAME_SIZE];
+};
+
+struct i40e_metadata_segment {
+ struct i40e_generic_seg_header header;
+ struct i40e_ppp_version version;
+ u32 track_id;
+ char name[I40E_PPP_NAME_SIZE];
+};
+
+struct i40e_device_id_entry {
+ u32 vendor_dev_id;
+ u32 sub_vendor_dev_id;
+};
+
+struct i40e_profile_segment {
+ struct i40e_generic_seg_header header;
+ struct i40e_ppp_version version;
+ char name[I40E_PPP_NAME_SIZE];
+ u32 device_table_count;
+ struct i40e_device_id_entry device_table[1];
+};
+
+struct i40e_section_table {
+ u32 section_count;
+ u32 section_offset[1];
+};
+
+struct i40e_profile_section_header {
+ u16 tbl_size;
+ u16 data_end;
+ struct {
+#define SECTION_TYPE_INFO 0x00000010
+#define SECTION_TYPE_MMIO 0x00000800
+#define SECTION_TYPE_AQ 0x00000801
+#define SECTION_TYPE_NOTE 0x80000000
+#define SECTION_TYPE_NAME 0x80000001
+ u32 type;
+ u32 offset;
+ u32 size;
+ } section;
+};
+
+struct i40e_profile_info {
+ u32 track_id;
+ struct i40e_ppp_version version;
+ u8 op;
+#define I40E_PPP_ADD_TRACKID 0x01
+#define I40E_PPP_REMOVE_TRACKID 0x02
+ u8 reserved[7];
+ u8 name[I40E_PPP_NAME_SIZE];
+};
#endif /* _I40E_TYPE_H_ */
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_adminq_cmd.h b/drivers/net/ethernet/intel/i40evf/i40e_adminq_cmd.h
index c28cb8f27243..91d8786d386d 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/i40evf/i40e_adminq_cmd.h
@@ -190,6 +190,10 @@ enum i40e_admin_queue_opc {
i40e_aqc_opc_add_mirror_rule = 0x0260,
i40e_aqc_opc_delete_mirror_rule = 0x0261,
+ /* Pipeline Personalization Profile */
+ i40e_aqc_opc_write_personalization_profile = 0x0270,
+ i40e_aqc_opc_get_personalization_profile_list = 0x0271,
+
/* DCB commands */
i40e_aqc_opc_dcb_ignore_pfc = 0x0301,
i40e_aqc_opc_dcb_updated = 0x0302,
@@ -1426,6 +1430,36 @@ struct i40e_aqc_add_delete_mirror_rule_completion {
I40E_CHECK_CMD_LENGTH(i40e_aqc_add_delete_mirror_rule_completion);
+/* Pipeline Personalization Profile */
+struct i40e_aqc_write_personalization_profile {
+ u8 flags;
+ u8 reserved[3];
+ __le32 profile_track_id;
+ __le32 addr_high;
+ __le32 addr_low;
+};
+
+I40E_CHECK_CMD_LENGTH(i40e_aqc_write_personalization_profile);
+
+struct i40e_aqc_write_ppp_resp {
+ __le32 error_offset;
+ __le32 error_info;
+ __le32 addr_high;
+ __le32 addr_low;
+};
+
+struct i40e_aqc_get_applied_profiles {
+ u8 flags;
+#define I40E_AQC_GET_PPP_GET_CONF 0x1
+#define I40E_AQC_GET_PPP_GET_RDPU_CONF 0x2
+ u8 rsv[3];
+ __le32 reserved;
+ __le32 addr_high;
+ __le32 addr_low;
+};
+
+I40E_CHECK_CMD_LENGTH(i40e_aqc_get_applied_profiles);
+
/* DCB 0x03xx*/
/* PFC Ignore (direct 0x0301)
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_common.c b/drivers/net/ethernet/intel/i40evf/i40e_common.c
index 626fbf1ead4d..43f10761f4ba 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_common.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_common.c
@@ -1131,3 +1131,215 @@ i40e_status i40e_vf_reset(struct i40e_hw *hw)
return i40e_aq_send_msg_to_pf(hw, I40E_VIRTCHNL_OP_RESET_VF,
0, NULL, 0, NULL);
}
+
+/**
+ * i40evf_aq_write_ppp - Write pipeline personalization profile (ppp)
+ * @hw: pointer to the hw struct
+ * @buff: command buffer (size in bytes = buff_size)
+ * @buff_size: buffer size in bytes
+ * @track_id: package tracking id
+ * @error_offset: returns error offset
+ * @error_info: returns error information
+ * @cmd_details: pointer to command details structure or NULL
+ **/
+enum
+i40e_status_code i40evf_aq_write_ppp(struct i40e_hw *hw, void *buff,
+ u16 buff_size, u32 track_id,
+ u32 *error_offset, u32 *error_info,
+ struct i40e_asq_cmd_details *cmd_details)
+{
+ struct i40e_aq_desc desc;
+ struct i40e_aqc_write_personalization_profile *cmd =
+ (struct i40e_aqc_write_personalization_profile *)
+ &desc.params.raw;
+ struct i40e_aqc_write_ppp_resp *resp;
+ i40e_status status;
+
+ i40evf_fill_default_direct_cmd_desc(&desc,
+ i40e_aqc_opc_write_personalization_profile);
+
+ desc.flags |= cpu_to_le16(I40E_AQ_FLAG_BUF | I40E_AQ_FLAG_RD);
+ if (buff_size > I40E_AQ_LARGE_BUF)
+ desc.flags |= cpu_to_le16((u16)I40E_AQ_FLAG_LB);
+
+ desc.datalen = cpu_to_le16(buff_size);
+
+ cmd->profile_track_id = cpu_to_le32(track_id);
+
+ status = i40evf_asq_send_command(hw, &desc, buff, buff_size, cmd_details);
+ if (!status) {
+ resp = (struct i40e_aqc_write_ppp_resp *)&desc.params.raw;
+ if (error_offset)
+ *error_offset = le32_to_cpu(resp->error_offset);
+ if (error_info)
+ *error_info = le32_to_cpu(resp->error_info);
+ }
+
+ return status;
+}
+
+/**
+ * i40evf_aq_get_ppp_list - Read pipeline personalization profile (ppp)
+ * @hw: pointer to the hw struct
+ * @buff: command buffer (size in bytes = buff_size)
+ * @buff_size: buffer size in bytes
+ * @cmd_details: pointer to command details structure or NULL
+ **/
+enum
+i40e_status_code i40evf_aq_get_ppp_list(struct i40e_hw *hw, void *buff,
+ u16 buff_size, u8 flags,
+ struct i40e_asq_cmd_details *cmd_details)
+{
+ struct i40e_aq_desc desc;
+ struct i40e_aqc_get_applied_profiles *cmd =
+ (struct i40e_aqc_get_applied_profiles *)&desc.params.raw;
+ i40e_status status;
+
+ i40evf_fill_default_direct_cmd_desc(&desc,
+ i40e_aqc_opc_get_personalization_profile_list);
+
+ desc.flags |= cpu_to_le16((u16)I40E_AQ_FLAG_BUF);
+ if (buff_size > I40E_AQ_LARGE_BUF)
+ desc.flags |= cpu_to_le16((u16)I40E_AQ_FLAG_LB);
+ desc.datalen = cpu_to_le16(buff_size);
+
+ cmd->flags = flags;
+
+ status = i40evf_asq_send_command(hw, &desc, buff, buff_size, cmd_details);
+
+ return status;
+}
+
+/**
+ * i40evf_find_segment_in_package
+ * @segment_type: the segment type to search for (i.e., SEGMENT_TYPE_I40E)
+ * @pkg_hdr: pointer to the package header to be searched
+ *
+ * This function searches a package file for a particular segment type. On
+ * success it returns a pointer to the segment header, otherwise it will
+ * return NULL.
+ **/
+struct i40e_generic_seg_header *
+i40evf_find_segment_in_package(u32 segment_type,
+ struct i40e_package_header *pkg_hdr)
+{
+ struct i40e_generic_seg_header *segment;
+ u32 i;
+
+ /* Search all package segments for the requested segment type */
+ for (i = 0; i < pkg_hdr->segment_count; i++) {
+ segment =
+ (struct i40e_generic_seg_header *)((u8 *)pkg_hdr +
+ pkg_hdr->segment_offset[i]);
+
+ if (segment->type == segment_type)
+ return segment;
+ }
+
+ return NULL;
+}
+
+/**
+ * i40evf_write_profile
+ * @hw: pointer to the hardware structure
+ * @profile: pointer to the profile segment of the package to be downloaded
+ * @track_id: package tracking id
+ *
+ * Handles the download of a complete package.
+ */
+enum i40e_status_code
+i40evf_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile,
+ u32 track_id)
+{
+ i40e_status status = 0;
+ struct i40e_section_table *sec_tbl;
+ struct i40e_profile_section_header *sec = NULL;
+ u32 dev_cnt;
+ u32 vendor_dev_id;
+ u32 *nvm;
+ u32 section_size = 0;
+ u32 offset = 0, info = 0;
+ u32 i;
+
+ if (!track_id) {
+ i40e_debug(hw, I40E_DEBUG_PACKAGE, "Track_id can't be 0.");
+ return I40E_NOT_SUPPORTED;
+ }
+
+ dev_cnt = profile->device_table_count;
+
+ for (i = 0; i < dev_cnt; i++) {
+ vendor_dev_id = profile->device_table[i].vendor_dev_id;
+ if ((vendor_dev_id >> 16) == PCI_VENDOR_ID_INTEL)
+ if (hw->device_id == (vendor_dev_id & 0xFFFF))
+ break;
+ }
+ if (i == dev_cnt) {
+ i40e_debug(hw, I40E_DEBUG_PACKAGE, "Device doesn't support PPP");
+ return I40E_ERR_DEVICE_NOT_SUPPORTED;
+ }
+
+ nvm = (u32 *)&profile->device_table[dev_cnt];
+ sec_tbl = (struct i40e_section_table *)&nvm[nvm[0] + 1];
+
+ for (i = 0; i < sec_tbl->section_count; i++) {
+ sec = (struct i40e_profile_section_header *)((u8 *)profile +
+ sec_tbl->section_offset[i]);
+
+ /* Skip 'AQ', 'note' and 'name' sections */
+ if (sec->section.type != SECTION_TYPE_MMIO)
+ continue;
+
+ section_size = sec->section.size +
+ sizeof(struct i40e_profile_section_header);
+
+ /* Write profile */
+ status = i40evf_aq_write_ppp(hw, (void *)sec, (u16)section_size,
+ track_id, &offset, &info, NULL);
+ if (status) {
+ i40e_debug(hw, I40E_DEBUG_PACKAGE,
+ "Failed to write profile: offset %d, info %d",
+ offset, info);
+ break;
+ }
+ }
+ return status;
+}
+
+/**
+ * i40evf_add_pinfo_to_list
+ * @hw: pointer to the hardware structure
+ * @profile: pointer to the profile segment of the package
+ * @profile_info_sec: buffer for information section
+ * @track_id: package tracking id
+ *
+ * Register a profile to the list of loaded profiles.
+ */
+enum i40e_status_code
+i40evf_add_pinfo_to_list(struct i40e_hw *hw,
+ struct i40e_profile_segment *profile,
+ u8 *profile_info_sec, u32 track_id)
+{
+ i40e_status status = 0;
+ struct i40e_profile_section_header *sec = NULL;
+ struct i40e_profile_info *pinfo;
+ u32 offset = 0, info = 0;
+
+ sec = (struct i40e_profile_section_header *)profile_info_sec;
+ sec->tbl_size = 1;
+ sec->data_end = sizeof(struct i40e_profile_section_header) +
+ sizeof(struct i40e_profile_info);
+ sec->section.type = SECTION_TYPE_INFO;
+ sec->section.offset = sizeof(struct i40e_profile_section_header);
+ sec->section.size = sizeof(struct i40e_profile_info);
+ pinfo = (struct i40e_profile_info *)(profile_info_sec +
+ sec->section.offset);
+ pinfo->track_id = track_id;
+ pinfo->version = profile->version;
+ pinfo->op = I40E_PPP_ADD_TRACKID;
+ memcpy(pinfo->name, profile->name, I40E_PPP_NAME_SIZE);
+
+ status = i40evf_aq_write_ppp(hw, (void *)sec, sec->data_end,
+ track_id, &offset, &info, NULL);
+ return status;
+}
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_prototype.h b/drivers/net/ethernet/intel/i40evf/i40e_prototype.h
index ba6c6bda0e22..741223d5d809 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_prototype.h
+++ b/drivers/net/ethernet/intel/i40evf/i40e_prototype.h
@@ -122,4 +122,21 @@ i40e_status i40e_write_phy_register(struct i40e_hw *hw, u8 page, u16 reg,
u8 i40e_get_phy_address(struct i40e_hw *hw, u8 dev_num);
i40e_status i40e_blink_phy_link_led(struct i40e_hw *hw,
u32 time, u32 interval);
+i40e_status i40evf_aq_write_ppp(struct i40e_hw *hw, void *buff,
+ u16 buff_size, u32 track_id,
+ u32 *error_offset, u32 *error_info,
+ struct i40e_asq_cmd_details *cmd_details);
+i40e_status i40evf_aq_get_ppp_list(struct i40e_hw *hw, void *buff,
+ u16 buff_size, u8 flags,
+ struct i40e_asq_cmd_details *cmd_details);
+struct i40e_generic_seg_header *
+i40evf_find_segment_in_package(u32 segment_type,
+ struct i40e_package_header *pkg_header);
+enum i40e_status_code
+i40evf_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *i40e_seg,
+ u32 track_id);
+enum i40e_status_code
+i40evf_add_pinfo_to_list(struct i40e_hw *hw,
+ struct i40e_profile_segment *profile,
+ u8 *profile_info_sec, u32 track_id);
#endif /* _I40E_PROTOTYPE_H_ */
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_type.h b/drivers/net/ethernet/intel/i40evf/i40e_type.h
index 16bb88084bb9..bde7f24af1c6 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_type.h
+++ b/drivers/net/ethernet/intel/i40evf/i40e_type.h
@@ -78,6 +78,7 @@ enum i40e_debug_mask {
I40E_DEBUG_DCB = 0x00000400,
I40E_DEBUG_DIAG = 0x00000800,
I40E_DEBUG_FD = 0x00001000,
+ I40E_DEBUG_PACKAGE = 0x00002000,
I40E_DEBUG_AQ_MESSAGE = 0x01000000,
I40E_DEBUG_AQ_DESCRIPTOR = 0x02000000,
@@ -1396,4 +1397,83 @@ enum i40e_reset_type {
#define I40E_FD_INSET_FLEX_WORD57_SHIFT 10
#define I40E_FD_INSET_FLEX_WORD57_MASK (0x1ULL << \
I40E_FD_INSET_FLEX_WORD57_SHIFT)
+
+/* Version format for PPP */
+struct i40e_ppp_version {
+ u8 major;
+ u8 minor;
+ u8 update;
+ u8 draft;
+};
+
+#define I40E_PPP_NAME_SIZE 32
+
+/* Package header */
+struct i40e_package_header {
+ struct i40e_ppp_version version;
+ u32 segment_count;
+ u32 segment_offset[1];
+};
+
+/* Generic segment header */
+struct i40e_generic_seg_header {
+#define SEGMENT_TYPE_METADATA 0x00000001
+#define SEGMENT_TYPE_NOTES 0x00000002
+#define SEGMENT_TYPE_I40E 0x00000011
+#define SEGMENT_TYPE_X722 0x00000012
+ u32 type;
+ struct i40e_ppp_version version;
+ u32 size;
+ char name[I40E_PPP_NAME_SIZE];
+};
+
+struct i40e_metadata_segment {
+ struct i40e_generic_seg_header header;
+ struct i40e_ppp_version version;
+ u32 track_id;
+ char name[I40E_PPP_NAME_SIZE];
+};
+
+struct i40e_device_id_entry {
+ u32 vendor_dev_id;
+ u32 sub_vendor_dev_id;
+};
+
+struct i40e_profile_segment {
+ struct i40e_generic_seg_header header;
+ struct i40e_ppp_version version;
+ char name[I40E_PPP_NAME_SIZE];
+ u32 device_table_count;
+ struct i40e_device_id_entry device_table[1];
+};
+
+struct i40e_section_table {
+ u32 section_count;
+ u32 section_offset[1];
+};
+
+struct i40e_profile_section_header {
+ u16 tbl_size;
+ u16 data_end;
+ struct {
+#define SECTION_TYPE_INFO 0x00000010
+#define SECTION_TYPE_MMIO 0x00000800
+#define SECTION_TYPE_AQ 0x00000801
+#define SECTION_TYPE_NOTE 0x80000000
+#define SECTION_TYPE_NAME 0x80000001
+ u32 type;
+ u32 offset;
+ u32 size;
+ } section;
+};
+
+struct i40e_profile_info {
+ u32 track_id;
+ struct i40e_ppp_version version;
+ u8 op;
+#define I40E_PPP_ADD_TRACKID 0x01
+#define I40E_PPP_REMOVE_TRACKID 0x02
+ u8 reserved[7];
+ u8 name[I40E_PPP_NAME_SIZE];
+};
#endif /* _I40E_TYPE_H_ */
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 07/14] i40e: don't hold RTNL lock while waiting for VF reset to finish
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (5 preceding siblings ...)
2017-04-20 1:57 ` [net-next 06/14] i40e: new AQ commands Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 08/14] i40e: factor out queue control from i40e_vsi_control_(tx|rx) Jeff Kirsher
` (7 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Jacob Keller, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Jacob Keller <jacob.e.keller@intel.com>
We made some effort to reduce the RTNL lock scope when resetting and
rebuilding the PF. Unfortunately we still held the RTNL lock during the
VF reset operation, which meant that multiple PFs could not reset in
parallel due to the global lock. For now, further reduce the scope by
not holding the RTNL lock while resetting VFs. This allows multiple PFs
to reset in a timely manner.
Change-ID: I2fbf823a0063f24dff67676cad09f0bbf83ee4ce
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e_main.c | 11 +++++++++--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 5e625e0c73ac..5c2ceb247959 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -7108,6 +7108,10 @@ static void i40e_rebuild(struct i40e_pf *pf, bool reinit, bool lock_acquired)
/* restart the VSIs that were rebuilt and running before the reset */
i40e_pf_unquiesce_all_vsi(pf);
+ /* Release the RTNL lock before we start resetting VFs */
+ if (!lock_acquired)
+ rtnl_unlock();
+
if (pf->num_alloc_vfs) {
for (v = 0; v < pf->num_alloc_vfs; v++)
i40e_reset_vf(&pf->vf[v], true);
@@ -7116,9 +7120,12 @@ static void i40e_rebuild(struct i40e_pf *pf, bool reinit, bool lock_acquired)
/* tell the firmware that we're starting */
i40e_send_version(pf);
+ /* We've already released the lock, so don't do it again */
+ goto end_core_reset;
+
end_unlock:
-if (!lock_acquired)
- rtnl_unlock();
+ if (!lock_acquired)
+ rtnl_unlock();
end_core_reset:
clear_bit(__I40E_RESET_FAILED, &pf->state);
clear_recovery:
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 08/14] i40e: factor out queue control from i40e_vsi_control_(tx|rx)
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (6 preceding siblings ...)
2017-04-20 1:57 ` [net-next 07/14] i40e: don't hold RTNL lock while waiting for VF reset to finish Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 09/14] i40e: fix CONFIG_BUSY checks in i40e_set_settings function Jeff Kirsher
` (6 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Jacob Keller, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Jacob Keller <jacob.e.keller@intel.com>
A future patch will need to be able to handle controlling queues without
waiting until all VSIs are handled. Factor out the direct queue
modification so that we can easily re-use this code. The result is also
a bit easier to read since we don't embed multiple single-letter loop
counters.
Change-ID: Id923cbfa43127b1c24d8ed4f809b1012c736d9ac
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e_main.c | 139 ++++++++++++++++++----------
1 file changed, 89 insertions(+), 50 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 5c2ceb247959..2bf5bb1b4627 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -3919,6 +3919,8 @@ static void i40e_netpoll(struct net_device *netdev)
}
#endif
+#define I40E_QTX_ENA_WAIT_COUNT 50
+
/**
* i40e_pf_txq_wait - Wait for a PF's Tx queue to be enabled or disabled
* @pf: the PF being configured
@@ -3949,6 +3951,50 @@ static int i40e_pf_txq_wait(struct i40e_pf *pf, int pf_q, bool enable)
}
/**
+ * i40e_control_tx_q - Start or stop a particular Tx queue
+ * @pf: the PF structure
+ * @pf_q: the PF queue to configure
+ * @enable: start or stop the queue
+ *
+ * This function enables or disables a single queue. Note that any delay
+ * required after the operation is expected to be handled by the caller of
+ * this function.
+ **/
+static void i40e_control_tx_q(struct i40e_pf *pf, int pf_q, bool enable)
+{
+ struct i40e_hw *hw = &pf->hw;
+ u32 tx_reg;
+ int i;
+
+ /* warn the TX unit of coming changes */
+ i40e_pre_tx_queue_cfg(&pf->hw, pf_q, enable);
+ if (!enable)
+ usleep_range(10, 20);
+
+ for (i = 0; i < I40E_QTX_ENA_WAIT_COUNT; i++) {
+ tx_reg = rd32(hw, I40E_QTX_ENA(pf_q));
+ if (((tx_reg >> I40E_QTX_ENA_QENA_REQ_SHIFT) & 1) ==
+ ((tx_reg >> I40E_QTX_ENA_QENA_STAT_SHIFT) & 1))
+ break;
+ usleep_range(1000, 2000);
+ }
+
+ /* Skip if the queue is already in the requested state */
+ if (enable == !!(tx_reg & I40E_QTX_ENA_QENA_STAT_MASK))
+ return;
+
+ /* turn on/off the queue */
+ if (enable) {
+ wr32(hw, I40E_QTX_HEAD(pf_q), 0);
+ tx_reg |= I40E_QTX_ENA_QENA_REQ_MASK;
+ } else {
+ tx_reg &= ~I40E_QTX_ENA_QENA_REQ_MASK;
+ }
+
+ wr32(hw, I40E_QTX_ENA(pf_q), tx_reg);
+}
+
+/**
* i40e_vsi_control_tx - Start or stop a VSI's rings
* @vsi: the VSI being configured
* @enable: start or stop the rings
@@ -3956,39 +4002,13 @@ static int i40e_pf_txq_wait(struct i40e_pf *pf, int pf_q, bool enable)
static int i40e_vsi_control_tx(struct i40e_vsi *vsi, bool enable)
{
struct i40e_pf *pf = vsi->back;
- struct i40e_hw *hw = &pf->hw;
- int i, j, pf_q, ret = 0;
- u32 tx_reg;
+ int i, pf_q, ret = 0;
pf_q = vsi->base_queue;
for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
+ i40e_control_tx_q(pf, pf_q, enable);
- /* warn the TX unit of coming changes */
- i40e_pre_tx_queue_cfg(&pf->hw, pf_q, enable);
- if (!enable)
- usleep_range(10, 20);
-
- for (j = 0; j < 50; j++) {
- tx_reg = rd32(hw, I40E_QTX_ENA(pf_q));
- if (((tx_reg >> I40E_QTX_ENA_QENA_REQ_SHIFT) & 1) ==
- ((tx_reg >> I40E_QTX_ENA_QENA_STAT_SHIFT) & 1))
- break;
- usleep_range(1000, 2000);
- }
- /* Skip if the queue is already in the requested state */
- if (enable == !!(tx_reg & I40E_QTX_ENA_QENA_STAT_MASK))
- continue;
-
- /* turn on/off the queue */
- if (enable) {
- wr32(hw, I40E_QTX_HEAD(pf_q), 0);
- tx_reg |= I40E_QTX_ENA_QENA_REQ_MASK;
- } else {
- tx_reg &= ~I40E_QTX_ENA_QENA_REQ_MASK;
- }
-
- wr32(hw, I40E_QTX_ENA(pf_q), tx_reg);
- /* No waiting for the Tx queue to disable */
+ /* Don't wait to disable when port Tx is suspended */
if (!enable && test_bit(__I40E_PORT_TX_SUSPENDED, &pf->state))
continue;
@@ -4035,6 +4055,43 @@ static int i40e_pf_rxq_wait(struct i40e_pf *pf, int pf_q, bool enable)
}
/**
+ * i40e_control_rx_q - Start or stop a particular Rx queue
+ * @pf: the PF structure
+ * @pf_q: the PF queue to configure
+ * @enable: start or stop the queue
+ *
+ * This function enables or disables a single queue. Note that any delay
+ * required after the operation is expected to be handled by the caller of
+ * this function.
+ **/
+static void i40e_control_rx_q(struct i40e_pf *pf, int pf_q, bool enable)
+{
+ struct i40e_hw *hw = &pf->hw;
+ u32 rx_reg;
+ int i;
+
+ for (i = 0; i < I40E_QTX_ENA_WAIT_COUNT; i++) {
+ rx_reg = rd32(hw, I40E_QRX_ENA(pf_q));
+ if (((rx_reg >> I40E_QRX_ENA_QENA_REQ_SHIFT) & 1) ==
+ ((rx_reg >> I40E_QRX_ENA_QENA_STAT_SHIFT) & 1))
+ break;
+ usleep_range(1000, 2000);
+ }
+
+ /* Skip if the queue is already in the requested state */
+ if (enable == !!(rx_reg & I40E_QRX_ENA_QENA_STAT_MASK))
+ return;
+
+ /* turn on/off the queue */
+ if (enable)
+ rx_reg |= I40E_QRX_ENA_QENA_REQ_MASK;
+ else
+ rx_reg &= ~I40E_QRX_ENA_QENA_REQ_MASK;
+
+ wr32(hw, I40E_QRX_ENA(pf_q), rx_reg);
+}
+
+/**
* i40e_vsi_control_rx - Start or stop a VSI's rings
* @vsi: the VSI being configured
* @enable: start or stop the rings
@@ -4042,31 +4099,13 @@ static int i40e_pf_rxq_wait(struct i40e_pf *pf, int pf_q, bool enable)
static int i40e_vsi_control_rx(struct i40e_vsi *vsi, bool enable)
{
struct i40e_pf *pf = vsi->back;
- struct i40e_hw *hw = &pf->hw;
- int i, j, pf_q, ret = 0;
- u32 rx_reg;
+ int i, pf_q, ret = 0;
pf_q = vsi->base_queue;
for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
- for (j = 0; j < 50; j++) {
- rx_reg = rd32(hw, I40E_QRX_ENA(pf_q));
- if (((rx_reg >> I40E_QRX_ENA_QENA_REQ_SHIFT) & 1) ==
- ((rx_reg >> I40E_QRX_ENA_QENA_STAT_SHIFT) & 1))
- break;
- usleep_range(1000, 2000);
- }
+ i40e_control_rx_q(pf, pf_q, enable);
- /* Skip if the queue is already in the requested state */
- if (enable == !!(rx_reg & I40E_QRX_ENA_QENA_STAT_MASK))
- continue;
-
- /* turn on/off the queue */
- if (enable)
- rx_reg |= I40E_QRX_ENA_QENA_REQ_MASK;
- else
- rx_reg &= ~I40E_QRX_ENA_QENA_REQ_MASK;
- wr32(hw, I40E_QRX_ENA(pf_q), rx_reg);
- /* No waiting for the Tx queue to disable */
+ /* Don't wait to disable when port Tx is suspended */
if (!enable && test_bit(__I40E_PORT_TX_SUSPENDED, &pf->state))
continue;
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 09/14] i40e: fix CONFIG_BUSY checks in i40e_set_settings function
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (7 preceding siblings ...)
2017-04-20 1:57 ` [net-next 08/14] i40e: factor out queue control from i40e_vsi_control_(tx|rx) Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 10/14] i40e: reduce wait time for adminq command completion Jeff Kirsher
` (5 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Jacob Keller, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Jacob Keller <jacob.e.keller@intel.com>
The check for I40E_CONFIG_BUSY state bit in the i40e_set_link_ksettings
function is fishy. First we can notice a few things about the check here.
First a similar check was introduced by commit
'c7d05ca89f8e ("i40e: driver ethtool core")'
Later a commit introducing the link settings was added by commit
'bf9c71417f72 ("i40e: Implement set_settings for ethtool")'
However, this second check was against vsi->state instead of pf->state,
and also failed to set the bit, it only checks. That indicates the locking
was not quite correct. The only other place that the state bit
in vsi->state gets used is to protect the filter list.
Since this code does not care about the mac filter list, and seems
clear the original code should have set the pf->state bit. Fix these
issues by using pf->state correctly, and by actually setting the bit
so that we properly lock as expected.
Since these checks occur while holding the rtnl_lock(), lets also add a
timeout so that we don't potentially softlock the system.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e_ethtool.c | 38 ++++++++++++++++++++------
1 file changed, 29 insertions(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
index 10325b5a9805..08035c4389cd 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
@@ -698,6 +698,7 @@ static int i40e_set_link_ksettings(struct net_device *netdev,
struct ethtool_link_ksettings copy_cmd;
i40e_status status = 0;
bool change = false;
+ int timeout = 50;
int err = 0;
u32 autoneg;
u32 advertise;
@@ -756,14 +757,20 @@ static int i40e_set_link_ksettings(struct net_device *netdev,
if (memcmp(©_cmd, &safe_cmd, sizeof(struct ethtool_link_ksettings)))
return -EOPNOTSUPP;
- while (test_bit(__I40E_CONFIG_BUSY, &vsi->state))
+ while (test_and_set_bit(__I40E_CONFIG_BUSY, &pf->state)) {
+ timeout--;
+ if (!timeout)
+ return -EBUSY;
usleep_range(1000, 2000);
+ }
/* Get the current phy config */
status = i40e_aq_get_phy_capabilities(hw, false, false, &abilities,
NULL);
- if (status)
- return -EAGAIN;
+ if (status) {
+ err = -EAGAIN;
+ goto done;
+ }
/* Copy abilities to config in case autoneg is not
* set below
@@ -779,7 +786,8 @@ static int i40e_set_link_ksettings(struct net_device *netdev,
if (!ethtool_link_ksettings_test_link_mode(
&safe_cmd, supported, Autoneg)) {
netdev_info(netdev, "Autoneg not supported on this phy\n");
- return -EINVAL;
+ err = -EINVAL;
+ goto done;
}
/* Autoneg is allowed to change */
config.abilities = abilities.abilities |
@@ -797,7 +805,8 @@ static int i40e_set_link_ksettings(struct net_device *netdev,
hw->phy.link_info.phy_type !=
I40E_PHY_TYPE_10GBASE_T) {
netdev_info(netdev, "Autoneg cannot be disabled on this phy\n");
- return -EINVAL;
+ err = -EINVAL;
+ goto done;
}
/* Autoneg is allowed to change */
config.abilities = abilities.abilities &
@@ -808,8 +817,10 @@ static int i40e_set_link_ksettings(struct net_device *netdev,
ethtool_convert_link_mode_to_legacy_u32(&tmp,
safe_cmd.link_modes.supported);
- if (advertise & ~tmp)
- return -EINVAL;
+ if (advertise & ~tmp) {
+ err = -EINVAL;
+ goto done;
+ }
if (advertise & ADVERTISED_100baseT_Full)
config.link_speed |= I40E_LINK_SPEED_100MB;
@@ -865,7 +876,8 @@ static int i40e_set_link_ksettings(struct net_device *netdev,
netdev_info(netdev, "Set phy config failed, err %s aq_err %s\n",
i40e_stat_str(hw, status),
i40e_aq_str(hw, hw->aq.asq_last_status));
- return -EAGAIN;
+ err = -EAGAIN;
+ goto done;
}
status = i40e_update_link_info(hw);
@@ -878,6 +890,9 @@ static int i40e_set_link_ksettings(struct net_device *netdev,
netdev_info(netdev, "Nothing changed, exiting without setting anything.\n");
}
+done:
+ clear_bit(__I40E_CONFIG_BUSY, &pf->state);
+
return err;
}
@@ -1292,6 +1307,7 @@ static int i40e_set_ringparam(struct net_device *netdev,
struct i40e_vsi *vsi = np->vsi;
struct i40e_pf *pf = vsi->back;
u32 new_rx_count, new_tx_count;
+ int timeout = 50;
int i, err = 0;
if ((ring->rx_mini_pending) || (ring->rx_jumbo_pending))
@@ -1316,8 +1332,12 @@ static int i40e_set_ringparam(struct net_device *netdev,
(new_rx_count == vsi->rx_rings[0]->count))
return 0;
- while (test_and_set_bit(__I40E_CONFIG_BUSY, &pf->state))
+ while (test_and_set_bit(__I40E_CONFIG_BUSY, &pf->state)) {
+ timeout--;
+ if (!timeout)
+ return -EBUSY;
usleep_range(1000, 2000);
+ }
if (!netif_running(vsi->netdev)) {
/* simple case - set for the next time the netdev is started */
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 10/14] i40e: reduce wait time for adminq command completion
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (8 preceding siblings ...)
2017-04-20 1:57 ` [net-next 09/14] i40e: fix CONFIG_BUSY checks in i40e_set_settings function Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 11/14] i40e: remove I40E_FLAG_IN_NETPOLL entirely Jeff Kirsher
` (4 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Jacob Keller, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Jacob Keller <jacob.e.keller@intel.com>
When sending an adminq command, we wait for the command to complete in
a loop. This loop waits for an entire millisecond, when in practice the
adminq command is processed often much faster.
Change the loop to use i40e_usec_delay instead, and wait for 50 usecs
each time instead. This appears to be about the minimum time required,
based on some manual observation and testing.
The primary benefit of this change is reducing latency of various
operations in the PF driver, especially when related to having a large
number of VFs enabled.
For example, on Linux, when instantiating 128 VFs, the time to finish
the operation dropped from about 9 seconds down to under 6 seconds.
Additionally, the time it takes to finish a PF reset with 128 VFs
dropped from 5.1 seconds down to 0.7 seconds.
As the examples above show, a significant portion of the delay is wasted
waiting for admiqn operations which have already finished.
This patch shouldn't cause impact to functionality, as we still check
and keep waiting until the command does get processed. The only expected
change is an increase in CPU utilization as we now check for completion
far more times. However, in practice the commands appear to generally be
complete within the first delay window anyways.
Change-ID: If8af8388e100da0a14eaf9e1af3afadf73a958cf
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e_adminq.c | 4 ++--
drivers/net/ethernet/intel/i40e/i40e_adminq.h | 2 +-
drivers/net/ethernet/intel/i40evf/i40e_adminq.c | 4 ++--
drivers/net/ethernet/intel/i40evf/i40e_adminq.h | 2 +-
4 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_adminq.c b/drivers/net/ethernet/intel/i40e/i40e_adminq.c
index 56fb27298936..ba04988e0598 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_adminq.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_adminq.c
@@ -850,8 +850,8 @@ i40e_status i40e_asq_send_command(struct i40e_hw *hw,
*/
if (i40e_asq_done(hw))
break;
- usleep_range(1000, 2000);
- total_delay++;
+ udelay(50);
+ total_delay += 50;
} while (total_delay < hw->aq.asq_cmd_timeout);
}
diff --git a/drivers/net/ethernet/intel/i40e/i40e_adminq.h b/drivers/net/ethernet/intel/i40e/i40e_adminq.h
index d92aad38afdc..2349fbe04bd2 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_adminq.h
+++ b/drivers/net/ethernet/intel/i40e/i40e_adminq.h
@@ -151,7 +151,7 @@ static inline int i40e_aq_rc_to_posix(int aq_ret, int aq_rc)
/* general information */
#define I40E_AQ_LARGE_BUF 512
-#define I40E_ASQ_CMD_TIMEOUT 250 /* msecs */
+#define I40E_ASQ_CMD_TIMEOUT 250000 /* usecs */
void i40e_fill_default_direct_cmd_desc(struct i40e_aq_desc *desc,
u16 opcode);
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_adminq.c b/drivers/net/ethernet/intel/i40evf/i40e_adminq.c
index 96385156b824..8b0d4b255dea 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_adminq.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_adminq.c
@@ -797,8 +797,8 @@ i40e_status i40evf_asq_send_command(struct i40e_hw *hw,
*/
if (i40evf_asq_done(hw))
break;
- usleep_range(1000, 2000);
- total_delay++;
+ udelay(50);
+ total_delay += 50;
} while (total_delay < hw->aq.asq_cmd_timeout);
}
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_adminq.h b/drivers/net/ethernet/intel/i40evf/i40e_adminq.h
index 1f9b3b5d946d..e0bfaa3d4a21 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_adminq.h
+++ b/drivers/net/ethernet/intel/i40evf/i40e_adminq.h
@@ -151,7 +151,7 @@ static inline int i40e_aq_rc_to_posix(int aq_ret, int aq_rc)
/* general information */
#define I40E_AQ_LARGE_BUF 512
-#define I40E_ASQ_CMD_TIMEOUT 250 /* msecs */
+#define I40E_ASQ_CMD_TIMEOUT 250000 /* usecs */
void i40evf_fill_default_direct_cmd_desc(struct i40e_aq_desc *desc,
u16 opcode);
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 11/14] i40e: remove I40E_FLAG_IN_NETPOLL entirely
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (9 preceding siblings ...)
2017-04-20 1:57 ` [net-next 10/14] i40e: reduce wait time for adminq command completion Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 12/14] i40e: split some code in i40e_reset_vf into helpers Jeff Kirsher
` (3 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Jacob Keller, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Jacob Keller <jacob.e.keller@intel.com>
This flag was originally intended to be used to let some
driver code know when we were running from netpoll.
Ultimately this was not necessary and we never used it.
Let's remove it
Change-ID: I43b72483d91c1638071d2a7f389ab171ec5b796a
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40evf/i40evf.h | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf.h b/drivers/net/ethernet/intel/i40evf/i40evf.h
index 7bbd11abed08..40f56e2335df 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf.h
+++ b/drivers/net/ethernet/intel/i40evf/i40evf.h
@@ -202,7 +202,6 @@ struct i40evf_adapter {
u32 flags;
#define I40EVF_FLAG_RX_CSUM_ENABLED BIT(0)
-#define I40EVF_FLAG_IN_NETPOLL BIT(4)
#define I40EVF_FLAG_IMIR_ENABLED BIT(5)
#define I40EVF_FLAG_MQ_CAPABLE BIT(6)
#define I40EVF_FLAG_PF_COMMS_FAILED BIT(8)
@@ -221,7 +220,6 @@ struct i40evf_adapter {
/* duplicates for common code */
#define I40E_FLAG_FDIR_ATR_ENABLED 0
#define I40E_FLAG_DCB_ENABLED 0
-#define I40E_FLAG_IN_NETPOLL I40EVF_FLAG_IN_NETPOLL
#define I40E_FLAG_RX_CSUM_ENABLED I40EVF_FLAG_RX_CSUM_ENABLED
#define I40E_FLAG_WB_ON_ITR_CAPABLE I40EVF_FLAG_WB_ON_ITR_CAPABLE
#define I40E_FLAG_OUTER_UDP_CSUM_CAPABLE I40EVF_FLAG_OUTER_UDP_CSUM_CAPABLE
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 12/14] i40e: split some code in i40e_reset_vf into helpers
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (10 preceding siblings ...)
2017-04-20 1:57 ` [net-next 11/14] i40e: remove I40E_FLAG_IN_NETPOLL entirely Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 13/14] i40e: reset all VFs in parallel when rebuilding PF Jeff Kirsher
` (2 subsequent siblings)
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Jacob Keller, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Jacob Keller <jacob.e.keller@intel.com>
A future patch is going to want to re-use some of the code in
i40e_reset_vf, so lets break up the beginning and ending parts into
their own helper functions. The first function will be used to
initialize the reset on a VF, while the second function will be used to
finalize the reset and restore functionality.
Change-ID: I48df808b8bf09de3c2ed8c521f57b3f0ab9e5907
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c | 109 ++++++++++++++-------
1 file changed, 72 insertions(+), 37 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
index 65c95ffc15ec..9a75ce3f1564 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
@@ -923,22 +923,19 @@ static int i40e_quiesce_vf_pci(struct i40e_vf *vf)
}
/**
- * i40e_reset_vf
+ * i40e_trigger_vf_reset
* @vf: pointer to the VF structure
* @flr: VFLR was issued or not
*
- * reset the VF
+ * Trigger hardware to start a reset for a particular VF. Expects the caller
+ * to wait the proper amount of time to allow hardware to reset the VF before
+ * it cleans up and restores VF functionality.
**/
-void i40e_reset_vf(struct i40e_vf *vf, bool flr)
+static void i40e_trigger_vf_reset(struct i40e_vf *vf, bool flr)
{
struct i40e_pf *pf = vf->pf;
struct i40e_hw *hw = &pf->hw;
u32 reg, reg_idx, bit_idx;
- bool rsd = false;
- int i;
-
- if (test_and_set_bit(__I40E_VF_DISABLE, &pf->state))
- return;
/* warn the VF */
clear_bit(I40E_VF_STAT_ACTIVE, &vf->vf_states);
@@ -970,37 +967,22 @@ void i40e_reset_vf(struct i40e_vf *vf, bool flr)
if (i40e_quiesce_vf_pci(vf))
dev_err(&pf->pdev->dev, "VF %d PCI transactions stuck\n",
vf->vf_id);
+}
- /* poll VPGEN_VFRSTAT reg to make sure
- * that reset is complete
- */
- for (i = 0; i < 10; i++) {
- /* VF reset requires driver to first reset the VF and then
- * poll the status register to make sure that the reset
- * completed successfully. Due to internal HW FIFO flushes,
- * we must wait 10ms before the register will be valid.
- */
- usleep_range(10000, 20000);
- reg = rd32(hw, I40E_VPGEN_VFRSTAT(vf->vf_id));
- if (reg & I40E_VPGEN_VFRSTAT_VFRD_MASK) {
- rsd = true;
- break;
- }
- }
-
- if (flr)
- usleep_range(10000, 20000);
-
- if (!rsd)
- dev_err(&pf->pdev->dev, "VF reset check timeout on VF %d\n",
- vf->vf_id);
-
- /* On initial reset, we won't have any queues */
- if (vf->lan_vsi_idx == 0)
- goto complete_reset;
+/**
+ * i40e_cleanup_reset_vf
+ * @vf: pointer to the VF structure
+ *
+ * Cleanup a VF after the hardware reset is finished. Expects the caller to
+ * have verified whether the reset is finished properly, and ensure the
+ * minimum amount of wait time has passed.
+ **/
+static void i40e_cleanup_reset_vf(struct i40e_vf *vf)
+{
+ struct i40e_pf *pf = vf->pf;
+ struct i40e_hw *hw = &pf->hw;
+ u32 reg;
- i40e_vsi_stop_rings(pf->vsi[vf->lan_vsi_idx]);
-complete_reset:
/* free VF resources to begin resetting the VSI state */
i40e_free_vf_res(vf);
@@ -1035,6 +1017,59 @@ void i40e_reset_vf(struct i40e_vf *vf, bool flr)
* request resources immediately after setting this flag.
*/
wr32(hw, I40E_VFGEN_RSTAT1(vf->vf_id), I40E_VFR_VFACTIVE);
+}
+
+/**
+ * i40e_reset_vf
+ * @vf: pointer to the VF structure
+ * @flr: VFLR was issued or not
+ *
+ * reset the VF
+ **/
+void i40e_reset_vf(struct i40e_vf *vf, bool flr)
+{
+ struct i40e_pf *pf = vf->pf;
+ struct i40e_hw *hw = &pf->hw;
+ bool rsd = false;
+ u32 reg;
+ int i;
+
+ /* If VFs have been disabled, there is no need to reset */
+ if (test_and_set_bit(__I40E_VF_DISABLE, &pf->state))
+ return;
+
+ i40e_trigger_vf_reset(vf, flr);
+
+ /* poll VPGEN_VFRSTAT reg to make sure
+ * that reset is complete
+ */
+ for (i = 0; i < 10; i++) {
+ /* VF reset requires driver to first reset the VF and then
+ * poll the status register to make sure that the reset
+ * completed successfully. Due to internal HW FIFO flushes,
+ * we must wait 10ms before the register will be valid.
+ */
+ usleep_range(10000, 20000);
+ reg = rd32(hw, I40E_VPGEN_VFRSTAT(vf->vf_id));
+ if (reg & I40E_VPGEN_VFRSTAT_VFRD_MASK) {
+ rsd = true;
+ break;
+ }
+ }
+
+ if (flr)
+ usleep_range(10000, 20000);
+
+ if (!rsd)
+ dev_err(&pf->pdev->dev, "VF reset check timeout on VF %d\n",
+ vf->vf_id);
+ usleep_range(10000, 20000);
+
+ /* On initial reset, we don't have any queues to disable */
+ if (vf->lan_vsi_idx != 0)
+ i40e_vsi_stop_rings(pf->vsi[vf->lan_vsi_idx]);
+
+ i40e_cleanup_reset_vf(vf);
i40e_flush(hw);
clear_bit(__I40E_VF_DISABLE, &pf->state);
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 13/14] i40e: reset all VFs in parallel when rebuilding PF
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (11 preceding siblings ...)
2017-04-20 1:57 ` [net-next 12/14] i40e: split some code in i40e_reset_vf into helpers Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-20 1:57 ` [net-next 14/14] i40e: use i40e_stop_rings_no_wait to implement PORT_SUSPENDED state Jeff Kirsher
2017-04-21 18:13 ` [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 David Miller
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Jacob Keller, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Jacob Keller <jacob.e.keller@intel.com>
When there are a lot of active VFs, it can take multiple seconds to
finish resetting all of them during certain flows., which can cause some
VFs to fail to wait long enough for the reset to occur. The user might
see messages like "Never saw reset" or "Reset never finished" and the VF
driver will stop functioning properly.
The naive solution would be to simply increase the wait timer. We can
get much more clever. Notice that i40e_reset_vf is run in a serialized
fashion, and includes lots of delays.
There are two prominent delays which take most of the time. First, when
we begin resetting VFs, we have multiple 10ms delays which accrue
because we reset each VF in a serial fashion. These delays accumulate to
almost 4 seconds when handling the maximum number of VFs (128).
Secondly, there is a massive 50ms delay for each time we disable queues
on a VSI. This delay is necessary to allow HW to finish disabling queues
before we restore functionality. However, just like with the first case,
we are paying the cost for each VF, rather than disabling all VFs and
waiting once.
Both of these can be fixed, but required some previous refactoring to
handle the special case. First, we will need the
i40e_vsi_wait_queues_disabled function which was previously DCB
specific. Second, we will need to implement our own
i40e_vsi_stop_rings_no_wait function which will handle the stopping of
rings without the delays.
Finally, implement an i40e_reset_all_vfs function, which will first
start the reset of all VFs, and pay the wait cost all at once, rather
than serially waiting for each VF before we start processing then next
one. After the VF has been reset, we'll disable all the VF queues, and
then wait for them to disable. Again, we'll organize the flow such that
we pay the wait cost only once.
Finally, after we've disabled queues we'll go ahead and begin restoring
VF functionality. The result is reducing the wait time by a large factor
and ensuring that VFs do not timeout when waiting in the VF driver.
Change-ID: Ia6e8cf8d98131b78aec89db78afb8d905c9b12be
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e.h | 2 +
drivers/net/ethernet/intel/i40e/i40e_main.c | 32 +++++--
drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c | 100 +++++++++++++++++++++
drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h | 1 +
4 files changed, 129 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index b629f3adcecb..dfab7b030b67 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -868,6 +868,8 @@ void i40e_notify_client_of_vf_msg(struct i40e_vsi *vsi, u32 vf_id,
int i40e_vsi_start_rings(struct i40e_vsi *vsi);
void i40e_vsi_stop_rings(struct i40e_vsi *vsi);
+void i40e_vsi_stop_rings_no_wait(struct i40e_vsi *vsi);
+int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi);
int i40e_reconfig_rss_queues(struct i40e_pf *pf, int queue_count);
struct i40e_veb *i40e_veb_setup(struct i40e_pf *pf, u16 flags, u16 uplink_seid,
u16 downlink_seid, u8 enabled_tc);
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 2bf5bb1b4627..2fe1fbdaafae 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -4159,6 +4159,29 @@ void i40e_vsi_stop_rings(struct i40e_vsi *vsi)
}
/**
+ * i40e_vsi_stop_rings_no_wait - Stop a VSI's rings and do not delay
+ * @vsi: the VSI being shutdown
+ *
+ * This function stops all the rings for a VSI but does not delay to verify
+ * that rings have been disabled. It is expected that the caller is shutting
+ * down multiple VSIs at once and will delay together for all the VSIs after
+ * initiating the shutdown. This is particularly useful for shutting down lots
+ * of VFs together. Otherwise, a large delay can be incurred while configuring
+ * each VSI in serial.
+ **/
+void i40e_vsi_stop_rings_no_wait(struct i40e_vsi *vsi)
+{
+ struct i40e_pf *pf = vsi->back;
+ int i, pf_q;
+
+ pf_q = vsi->base_queue;
+ for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
+ i40e_control_tx_q(pf, pf_q, false);
+ i40e_control_rx_q(pf, pf_q, false);
+ }
+}
+
+/**
* i40e_vsi_free_irq - Free the irq association with the OS
* @vsi: the VSI being configured
**/
@@ -4488,14 +4511,13 @@ static void i40e_pf_unquiesce_all_vsi(struct i40e_pf *pf)
}
}
-#ifdef CONFIG_I40E_DCB
/**
* i40e_vsi_wait_queues_disabled - Wait for VSI's queues to be disabled
* @vsi: the VSI being configured
*
* Wait until all queues on a given VSI have been disabled.
**/
-static int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
+int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
{
struct i40e_pf *pf = vsi->back;
int i, pf_q, ret;
@@ -4523,6 +4545,7 @@ static int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
return 0;
}
+#ifdef CONFIG_I40E_DCB
/**
* i40e_pf_wait_queues_disabled - Wait for all queues of PF VSIs to be disabled
* @pf: the PF
@@ -7151,10 +7174,7 @@ static void i40e_rebuild(struct i40e_pf *pf, bool reinit, bool lock_acquired)
if (!lock_acquired)
rtnl_unlock();
- if (pf->num_alloc_vfs) {
- for (v = 0; v < pf->num_alloc_vfs; v++)
- i40e_reset_vf(&pf->vf[v], true);
- }
+ i40e_reset_all_vfs(pf, true);
/* tell the firmware that we're starting */
i40e_send_version(pf);
diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
index 9a75ce3f1564..350cba70490c 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
@@ -1076,6 +1076,106 @@ void i40e_reset_vf(struct i40e_vf *vf, bool flr)
}
/**
+ * i40e_reset_all_vfs
+ * @pf: pointer to the PF structure
+ * @flr: VFLR was issued or not
+ *
+ * Reset all allocated VFs in one go. First, tell the hardware to reset each
+ * VF, then do all the waiting in one chunk, and finally finish restoring each
+ * VF after the wait. This is useful during PF routines which need to reset
+ * all VFs, as otherwise it must perform these resets in a serialized fashion.
+ **/
+void i40e_reset_all_vfs(struct i40e_pf *pf, bool flr)
+{
+ struct i40e_hw *hw = &pf->hw;
+ struct i40e_vf *vf;
+ int i, v;
+ u32 reg;
+
+ /* If we don't have any VFs, then there is nothing to reset */
+ if (!pf->num_alloc_vfs)
+ return;
+
+ /* If VFs have been disabled, there is no need to reset */
+ if (test_and_set_bit(__I40E_VF_DISABLE, &pf->state))
+ return;
+
+ /* Begin reset on all VFs at once */
+ for (v = 0; v < pf->num_alloc_vfs; v++)
+ i40e_trigger_vf_reset(&pf->vf[v], flr);
+
+ /* HW requires some time to make sure it can flush the FIFO for a VF
+ * when it resets it. Poll the VPGEN_VFRSTAT register for each VF in
+ * sequence to make sure that it has completed. We'll keep track of
+ * the VFs using a simple iterator that increments once that VF has
+ * finished resetting.
+ */
+ for (i = 0, v = 0; i < 10 && v < pf->num_alloc_vfs; i++) {
+ usleep_range(10000, 20000);
+
+ /* Check each VF in sequence, beginning with the VF to fail
+ * the previous check.
+ */
+ while (v < pf->num_alloc_vfs) {
+ vf = &pf->vf[v];
+ reg = rd32(hw, I40E_VPGEN_VFRSTAT(vf->vf_id));
+ if (!(reg & I40E_VPGEN_VFRSTAT_VFRD_MASK))
+ break;
+
+ /* If the current VF has finished resetting, move on
+ * to the next VF in sequence.
+ */
+ v++;
+ }
+ }
+
+ if (flr)
+ usleep_range(10000, 20000);
+
+ /* Display a warning if at least one VF didn't manage to reset in
+ * time, but continue on with the operation.
+ */
+ if (v < pf->num_alloc_vfs)
+ dev_err(&pf->pdev->dev, "VF reset check timeout on VF %d\n",
+ pf->vf[v].vf_id);
+ usleep_range(10000, 20000);
+
+ /* Begin disabling all the rings associated with VFs, but do not wait
+ * between each VF.
+ */
+ for (v = 0; v < pf->num_alloc_vfs; v++) {
+ /* On initial reset, we don't have any queues to disable */
+ if (pf->vf[v].lan_vsi_idx == 0)
+ continue;
+
+ i40e_vsi_stop_rings_no_wait(pf->vsi[pf->vf[v].lan_vsi_idx]);
+ }
+
+ /* Now that we've notified HW to disable all of the VF rings, wait
+ * until they finish.
+ */
+ for (v = 0; v < pf->num_alloc_vfs; v++) {
+ /* On initial reset, we don't have any queues to disable */
+ if (pf->vf[v].lan_vsi_idx == 0)
+ continue;
+
+ i40e_vsi_wait_queues_disabled(pf->vsi[pf->vf[v].lan_vsi_idx]);
+ }
+
+ /* Hw may need up to 50ms to finish disabling the RX queues. We
+ * minimize the wait by delaying only once for all VFs.
+ */
+ mdelay(50);
+
+ /* Finish the reset on each VF */
+ for (v = 0; v < pf->num_alloc_vfs; v++)
+ i40e_cleanup_reset_vf(&pf->vf[v]);
+
+ i40e_flush(hw);
+ clear_bit(__I40E_VF_DISABLE, &pf->state);
+}
+
+/**
* i40e_free_vfs
* @pf: pointer to the PF structure
*
diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h
index 37af437daa5d..9495f1422122 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h
+++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h
@@ -124,6 +124,7 @@ int i40e_vc_process_vf_msg(struct i40e_pf *pf, s16 vf_id, u32 v_opcode,
u32 v_retval, u8 *msg, u16 msglen);
int i40e_vc_process_vflr_event(struct i40e_pf *pf);
void i40e_reset_vf(struct i40e_vf *vf, bool flr);
+void i40e_reset_all_vfs(struct i40e_pf *pf, bool flr);
void i40e_vc_notify_vf_reset(struct i40e_vf *vf);
/* VF configuration related iplink handlers */
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* [net-next 14/14] i40e: use i40e_stop_rings_no_wait to implement PORT_SUSPENDED state
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (12 preceding siblings ...)
2017-04-20 1:57 ` [net-next 13/14] i40e: reset all VFs in parallel when rebuilding PF Jeff Kirsher
@ 2017-04-20 1:57 ` Jeff Kirsher
2017-04-21 18:13 ` [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 David Miller
14 siblings, 0 replies; 31+ messages in thread
From: Jeff Kirsher @ 2017-04-20 1:57 UTC (permalink / raw)
To: davem; +Cc: Jacob Keller, netdev, nhorman, sassmann, jogreene, Jeff Kirsher
From: Jacob Keller <jacob.e.keller@intel.com>
This state bit was added as a way for DCB to avoid having to wait for
the queues to disable when handling LLDP events. The logic for this was
burried deep within stop Tx and stop Rx queue code. First, let's rename
it so that it does not appear to only affect Tx when infact it modifies
both Tx and Rx flow. Second we can move it up into the i40e_stop_rings()
function, and we can simply re-use the i40e_stop_rings_no_wait() so that
we don't have to bury the implementation as deep into the call stack.
An alternative might be to remove the state bit and instead attempt to
shut down everything directly in DCP flow. This, however, is not ideal
because it creates yet another separate shutdown routine that we'd have
to maintain. In the current implementation any changes will be made to
both flows.
Change-ID: I68e1ccb901af320862bca395e9c9746f08e8b17c
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e.h | 2 +-
drivers/net/ethernet/intel/i40e/i40e_main.c | 16 ++++++----------
2 files changed, 7 insertions(+), 11 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index dfab7b030b67..70f9458f7a01 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -145,7 +145,7 @@ enum i40e_state_t {
__I40E_DOWN_REQUESTED,
__I40E_FD_FLUSH_REQUESTED,
__I40E_RESET_FAILED,
- __I40E_PORT_TX_SUSPENDED,
+ __I40E_PORT_SUSPENDED,
__I40E_VF_DISABLE,
};
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 2fe1fbdaafae..c001562f19b2 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -4008,10 +4008,6 @@ static int i40e_vsi_control_tx(struct i40e_vsi *vsi, bool enable)
for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
i40e_control_tx_q(pf, pf_q, enable);
- /* Don't wait to disable when port Tx is suspended */
- if (!enable && test_bit(__I40E_PORT_TX_SUSPENDED, &pf->state))
- continue;
-
/* wait for the change to finish */
ret = i40e_pf_txq_wait(pf, pf_q, enable);
if (ret) {
@@ -4105,10 +4101,6 @@ static int i40e_vsi_control_rx(struct i40e_vsi *vsi, bool enable)
for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
i40e_control_rx_q(pf, pf_q, enable);
- /* Don't wait to disable when port Tx is suspended */
- if (!enable && test_bit(__I40E_PORT_TX_SUSPENDED, &pf->state))
- continue;
-
/* wait for the change to finish */
ret = i40e_pf_rxq_wait(pf, pf_q, enable);
if (ret) {
@@ -4151,6 +4143,10 @@ int i40e_vsi_start_rings(struct i40e_vsi *vsi)
**/
void i40e_vsi_stop_rings(struct i40e_vsi *vsi)
{
+ /* When port TX is suspended, don't wait */
+ if (test_bit(__I40E_PORT_SUSPENDED, &vsi->back->state))
+ return i40e_vsi_stop_rings_no_wait(vsi);
+
/* do rx first for enable and last for disable
* Ignore return value, we need to shutdown whatever we can
*/
@@ -5948,7 +5944,7 @@ static int i40e_handle_lldp_event(struct i40e_pf *pf,
else
pf->flags &= ~I40E_FLAG_DCB_ENABLED;
- set_bit(__I40E_PORT_TX_SUSPENDED, &pf->state);
+ set_bit(__I40E_PORT_SUSPENDED, &pf->state);
/* Reconfiguration needed quiesce all VSIs */
i40e_pf_quiesce_all_vsi(pf);
@@ -5957,7 +5953,7 @@ static int i40e_handle_lldp_event(struct i40e_pf *pf,
ret = i40e_resume_port_tx(pf);
- clear_bit(__I40E_PORT_TX_SUSPENDED, &pf->state);
+ clear_bit(__I40E_PORT_SUSPENDED, &pf->state);
/* In case of error no point in resuming VSIs */
if (ret)
goto exit;
--
2.12.2
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19
2017-04-20 1:57 [net-next 00/14][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-19 Jeff Kirsher
` (13 preceding siblings ...)
2017-04-20 1:57 ` [net-next 14/14] i40e: use i40e_stop_rings_no_wait to implement PORT_SUSPENDED state Jeff Kirsher
@ 2017-04-21 18:13 ` David Miller
14 siblings, 0 replies; 31+ messages in thread
From: David Miller @ 2017-04-21 18:13 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, nhorman, sassmann, jogreene
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Wed, 19 Apr 2017 18:57:36 -0700
> This series contains updates to i40e and i40evf only, most notable being
> the addition of trace points for BPF programs.
Pulled, thanks Jeff.
^ permalink raw reply [flat|nested] 31+ messages in thread